Peak Memory

Follow this topic Ask about this topic

Momentum

54 papers in the last four weeks, up 157% on the four weeks before. 0.5% of all new papers.

Jul 13Week of Sep 28

Latest papers 279

Jun 8, 2026cs.LG

Scaling Neural Network Verification with Tensor Parallelism and Fully Sharded Data Parallelism

Formal neural network verification -- proving that a network satisfies safety properties for all inputs in a specified domain -- is bounded in practice by GPU memory: standard implementations of bound-propagation algorithms (IBP, CROWN,

α

-CROWN) require weight and relaxation-coefficient matrices to reside entirely on one accelerator. We adapt two parallelism techniques originally developed for large-scale model training to the auto_LiRPA /

α,β

-CROWN verification framework. Tensor Parallelism (TP) shards both weight and

A

-matrices across GPUs, achieving

{\approx}2\times

peak-memory reduction at

P{=}2

; soundness is confirmed on VNN-COMP 2022 MNIST-FC benchmarks, though bound tightness degrades with the number of sharded zones due to forced IBP substitution for intermediate bounds inside sharded zones. Fully Sharded Data Parallelism (FSDP) shards only weight matrices with a per-layer AllGather, producing bounds that are bitwise identical to the single-GPU baseline: baseline memory drops by 80--90%, peak memory by 34--39% on wide MLPs. FSDP integrates cleanly with complete verification (

β

-CROWN + Branch-and-Bound) and with convolutional layers (BoundConv); a complete unsat result is obtained for CIFAR-100 ResNet-large (VNN-COMP 2024) under FSDP. Across all experiments the memory bottleneck in

α

-CROWN+BaB mode proves to be per-neuron alpha tensors, not weight matrices, pointing to the key direction for future work.

Sergei Vorobyov, Eugene Ilyushin

Lomonosov Moscow State University · Central University

Neural Network Verification Cpu-Gpu Hybrid Designs Peak Memory

Jun 5, 2026cs.RO

Planning-aligned Token Compression for Long-Context Autonomous Driving

Monolithic vision-action models represent an emerging paradigm in autonomous driving. However, this architecture produces token sequences that quickly exceed real-time computational budgets when encoding extended temporal context for complex interactions. While approaches like linear transformers and external memory try to make the context lightweight, token compression is most compatible with the architecture as it requires no backbone modifications. Yet existing compression adopts rule-based heuristics like temporal decay, decoupled from planning, risking loss of decision-critical information. We propose COMPACT-VA, a planning-aligned working memory framework built on conditional VQ-VAE, compressing extended context into bounded representations. Compression is conditioned on both historical trajectory and a learned planning intent that the posterior encoder distills from future trajectories during training, while the prior encoder learns to predict it from compressed observations. The compressed memory, concatenated with the predicted latent, feeds the policy for end-to-end optimization, planning with retained decision-critical information. We evaluate on high-signal dynamic scenarios where historical context is most critical for behavior correctness (e.g., stop, yield, or proceed), and accordingly design behavioral metrics. Under comparable token budgets, we achieve

>

6% improvement (68.3%) on success rates with consistent gains across metrics. Ablations validate planning-aligned coupling effectiveness. Closed-loop evaluation confirms that COMPACT-VA maintained general driving performance with 3.3* speedup and 2.7* memory reduction over uncompressed processing.

Zhixuan Liang, Yuxiao Chen, Yurong You +12

NVIDIA Research · School of Computing and Data Science, The University of Hong Kong

Token Compression Autonomous Driving Diffusion-Based Vision-Language-Actions Data Compression Methods Peak Memory

Jun 5, 2026cs.AI

AdMem: Advanced Memory for Task-solving Agents

Large Language Models (LLMs) show promise as tool-using agents but remain limited in long-horizon tasks that require remembering, organizing, and reusing knowledge. Prior memory approaches aim to resolve the situation, but mainly focus on storing factual information. Recent work on procedural memory improves task reuse, yet often reduces to replaying past successes without addressing failure cases or online scalability. We introduce a unified and automatic memory framework that integrates semantic, episodic, and procedural memory in a bi-level design combining short-term and long-term stores. A multi-agent architecture with actor, memory, and critic agents enables automatic memory generation, reward annotation, and adaptive retrieval. Long-term memory is managed through reward-based evaluation, merging, and pruning, ensuring scalability and continual improvement. Experiments across various environments show that our approach improves robustness and success on long multi-turn tasks compared to existing baselines. This work highlights the importance of comprehensive, adaptive memory for advancing LLM-based agents.

Runzhe Wang, Huilin Lu, Shengjie Liu +2

Princeton University · Amazon · Arm

Long-Term Memory Semantic Memory Peak Memory Reward Functions

Jun 4, 2026cs.AI

Beyond Semantic Organization: Memory as Execution State Management for Long-Horizon Agents

LLM-based agents increasingly tackle long-horizon tasks with interdependent decisions, where each action reshapes future constraints and intermediate errors can cascade. Existing RAG and agent memory systems organize histories by semantic similarity, retrieving content-relevant entries at decision time. We argue that this design mismatches execution-state dependencies: it fragments decision trajectories and mixes valid and erroneous traces, hindering coherent state reconstruction and error isolation. We propose MAGE (Memory as Agent-Guided Exploration), an active execution-state manager that stores interactions in a hierarchical state tree. The agent derives its state from the active root-to-current path, combining subgoal summaries, recent traces, and hints from prior branches. Four coupled operations maintain the tree: Grow records new traces, Compress summarizes completed subgoals, Maintain validates summaries, and Revise restores a target boundary and resumes on a new branch. This design bounds context growth while preserving state integrity and isolating flawed segments from the active path. Experiments on MemoryArena show that MAGE improves the average task success rate by 7.8--20.4 pp over baselines, while reducing token consumption by 55.1%.

Yaoqi Chen, Haibin Lai, Yuru Feng +12

University of Science and Technology of China · Microsoft · University of California, San Diego +1

Long-Term Agent Memory Long-Horizon Agents Long-Horizon Task Planning Peak Memory

Jun 4, 2026cs.AI

When Should Memory Stay Silent: Measuring Memory-Use Boundaries in Memory-Augmented Conversational Agents

Long-term memory enables language model agents to support personalized interactions, but it remains unclear when available memories warrant integration into responses. Existing memory evaluations emphasize retrieval accuracy and downstream task utility, while overlooking whether retrieved sensitive memory content is warranted in the current turn. We introduce RBI-Eval, a controlled measurement study built around a probe set that compares model behavior with and without access to sensitive memory under identical benign prompts. We evaluate four base LLMs against a matched no-memory reference across four memory-access settings: full-context exposure and three retrieval systems. Our results reveal substantial behavioral divergence. With memory available, the separation score for sensitive-memory integration decreases by 8.9%--26.6% relative to the matched no-memory reference for GPT-5.4-mini, but by 51.1%--82.9% for Claude-Sonnet-4.6, DeepSeek-V4-Flash, and Qwen3.5-9B. Control experiments on DeepSeek and GPT-5.4-mini show this effect is specific to sensitive content, rather than general personalization. Retrieval systems reduce exposure but do not eliminate integration once sensitive memory reaches the generator. These findings suggest safe personalization requires memory-aware decisions at both retrieval and generation time.

Lingxiang Xu, Jiaoyun Yang, Min Hu +2

†Hefei University of Technology · ‡Harvard Medical School

Long-Term Memory Peak Memory Personalization

Jun 4, 2026cs.CV

MemoryCard: Topic-Aware Multi-Modal Clue Compression for Long-Video Question Answering

Long-video question answering remains challenging for Vision-Language Models (VLMs), as answer-relevant evidence is often sparse, transient, and temporally dispersed across lengthy video contexts. Existing frame-centric approaches improve efficiency through uniform sampling, query-aware frame selection, visual-token compression, and adaptive resolution strategies. However, they still rely on isolated and fragmented frames as the fundamental evidence units, limiting VLMs' ability to effectively capture coherent event-level semantics. To address this limitation, we propose MemoryCard, a video-memory-based augmentation framework that organizes long videos into self-contained Memory Cards. Specifically, MemoryCard first performs a self-reading process over videos and aligned utterances to segment the video into semantically coherent units, each corresponding to a distinct topic or event. For each unit, it generates an event-level video gist and selects representative visual moments, which are then rendered into unified Memory Cards for retrieval and question answering. Experimental results demonstrate that MemoryCard consistently improves long-video QA performance under comparable visual-token budgets, achieving up to a 21.8% relative improvement in accuracy. All code is available at https://github.com/NEUIR/MemoryCard.

Qing Yang, Pengcheng Huang, Xinze Li +6

School of Computer Science and Engineering, Northeastern University, Shenyang, China · Department of Computer Science and Technology, Tsinghua University, Beijing, China · Digital China Group, Beijing, China

Long Video Question Answering Peak Memory Data Compression Methods

Jun 3, 2026cs.CV

GOPAgen: Motion-Aware and Efficient Agentic Long-Video Understanding with Structural Memory and Hierarchical Reasoning

Despite significant progress in agentic long video understanding, existing methods still lack detailed motion comprehension coupled with an efficient memory architecture. In this paper, we propose GOPAgen, a novel approach that first integrates video codec into the video understanding framework via a meticulously designed motion agent trained on Groups of Pictures (GOPs) from video codec. We further develop a GOP tree reasoning algorithm, which is naturally aligned with video codec and enhances the model's ability to understand local detailed motions in videos. Additionally, we carefully design a structural memory mechanism that integrates local motion information with detailed captions in structural pages, and propose an efficient coarse-to-fine zoom-in algorithm to fully exploit the structural memory. Furthermore, we incorporate a motion vector database into the framework to enable efficient retrieval of motion vectors at different granularities. Overall, our method achieves superior Video Question Answering (VQA) performance on various video understanding benchmarks, including MotionBench and Egoschema, thereby demonstrating the superiority of our proposed framework.

Haozhe Chi, Yang Jin, Yadong Mu

Peking University

Long Video Question Answering Motion-Based Perception Peak Memory

Jun 3, 2026cs.LG

Operator Fusion for LLM Inference on the Tensix Architecture

This study addresses on-device inference bottlenecks of Transformer models on Tenstorrent's Tensix architecture and proposes an operator fusion strategy that enhances data locality. RMSNorm is fused with matrix multiplication in self-attention and in the FFN, enabling back-to-back execution of memory-bound and compute-bound operators in on-chip SRAM to significantly reduce DRAM reads/writes of intermediate results and scheduling overhead. To support multi-core parallelism, a NoC-based multicast mechanism is leveraged in which row/column master nodes efficiently distribute inputs and weights across the core mesh, alleviating DRAM bandwidth contention. Experiments on the Wormhole platform with Qwen2.5-0.5B, Qwen3-0.6B, and Qwen3-4B show up to 37.44% latency reduction for attention and 15.89% for MLP, with up to 7.91% reduction per decoder layer, while Pearson Correlation Coefficient (PCC) remains above 98.75%, confirming significant end-to-end efficiency gains under numerical consistency.

Qingbo Wu, Ke Li, Wenzhu Wang +3

KylinSoft Co., Ltd, Tianjin, China · openKylin Community, Tianjin, China · College of Computer Science and Technology, National University of Defense Technology, Changsha, China

Transformer Architectures LLM Inference Optimization Peak Memory Chip Design

Jun 2, 2026cs.CL

HybridThinker: Efficient Chain-of-Thought Reasoning via Compressed Memory and Transient Thought Steps

Extended chain-of-thought (CoT) traces improve LLM reasoning but incur substantial computational and memory costs. While existing CoT compression methods mitigate this by condensing thought steps into compact representations via memory tokens and retaining only these representations at inference time, the loss of fine-grained information makes subsequent steps more error-prone. To alleviate this, we propose \textbf{HybridThinker}, where in addition to preserved these representations, thought steps are also temporarily retained to provide fine-grained details. However, we observe that naively keeping thought steps accessible to subsequent steps \emph{during training} lets the model bypass memory tokens by retrieving information directly from these steps, leaving the model's ability to compress and retrieve information through memory tokens insufficiently trained. We therefore introduce a hybrid training scheme, in which only some thought steps are directly accessible through attention to subsequent steps, while the other thought steps are masked, forcing the model to use memory tokens for compression and retrieval. Across 4 reasoning benchmarks, HybridThinker matches the uncompressed baseline, advancing the state of the art in CoT compression by 5.8 points on average accuracy with similar inference time. Ablation studies confirm that both temporary thought-step retention and the hybrid training scheme contribute to these gains.

Xin Liu, Runsong Zhao, Xinyu Liu +8

School of Computer Science and Engineering, Northeastern University, Shenyang 110819, China · China Unicom Cloud-Link, China · NiuTrans Research, Shenyang, China

Chain-of-Thought Reasoning Latent Thoughts Peak Memory

Jun 1, 2026cs.AI

Evaluation of Baseline Methods for IDD-based SSD External Memory Search

Many difficult search problems cannot be solved by algorithms such as A* using only RAM. Search algorithms which use external memory such as SSDs and HDDs with much higher capacity than RAM have been proposed in previous work, but previous work has focused on delayed duplicate detection approaches, as well as complex immediate duplicate detection (IDD) methods, and relatively simple methods for IDD have not been systematically studied. In addition, the effect of OS-level mechanisms for managing and speeding up accesses to external memory, such as page caches, has not been studied. This paper addresses these gaps in the literature by evaluating and analyzing the performance of simple baseline approaches for IDD-based A*.

Yuki Suzuki, Alex Fukunaga

The University of Tokyo

Search Algorithms Peak Memory Cache

Jun 1, 2026cs.DC

Observation, Not Prediction: Conversation-Level Disaggregated Scheduling for Agentic Serving

LLM-based agents resolve a user task through many turns of dependent inference and tool calls, producing a workload whose total cost is unknown when the task arrives. Existing multi-turn systems keep the turn as the scheduling unit and decide, turn by turn, whether to disaggregate prefill from decode. That decision rests on the turn's decode length, tool behavior, and KV growth, quantities that are not observable when the scheduler must act, forcing the system to predict them. We show this dependence on prediction is imposed by the scheduling unit, not the workload. Raising the scheduling unit from the turn to the conversation converts turn-level irregularity into a stable, two-phase structure: 1) a compute-bound turn-1 prefill followed by 2) a long, memory-bound tail. Thus, with the conversation as the scheduling unit, placement reduces to reading the first-turn input length and per-decoder KV occupancy, both directly observable. We instantiate this principle in ConServe, which routes the first-turn prefill to a high-throughput prefiller, transfers the KV cache exactly once, and pins the conversation to a single decoder for its entire tail, with no learned model of decode-side cost. Against a per-turn prediction baseline, ConServe reduces p95 time-to-first-effective-token (the latency of a conversation's first user-visible output) by 51.08% and improves energy efficiency by 7.51% while preserving last-turn TBT and SLOs; mapping the two phases onto heterogeneous GPU tiers adds a further 22.75% in energy efficiency.

Jianru Ding, Ryien Hosseini, Pouya Mahdi Gholami +2

University of Chicago

Time-To-First-Token Large Language Model Serving Schedule Prefill Agentic Peak Memory

May 31, 2026cs.LG

Soft-NBCE: Entropy-Weighted Chunk Fusion for Long-Context

The quadratic complexity of self-attention remains a bottleneck for Large Language Models (LLMs) processing ultra-long contexts. The Naive Bayes Cognitive Engine (NBCE) parallelizes long-context inference by chunking documents and routing to the lowest-entropy chunk at each decoding step. This hard-selection strategy causes semantic fragmentation during cross-chunk reasoning, as abrupt routing changes between adjacent tokens disrupt the model's contextual grounding. We present Soft-NBCE, a lightweight extension that replaces discrete chunk selection with soft entropy-weighted chunk fusion. A temperature-scaled Softmax over predictive entropies assigns continuous weights to all chunks, enabling log-space aggregation across chunk-conditioned distributions. To partially compensate for the conditional independence assumption introduced by chunking, we propose Consistency Distillation, a LoRA-based self-distillation that constrains the chunked logit distribution toward a full-context teacher via KL-divergence. On LongBench multi-hop benchmarks, Soft-NBCE with Consistency Distillation improves consistently over NBCE-style baselines (MuSiQue F1: 0.310 vs.\ 0.275 for Vanilla NBCE; HotpotQA F1: 0.479 vs.\ 0.427) while maintaining retrieval accuracy (NIAH-32K: 0.909) at O(L^2/n) peak memory.

Shihao Ji, Mingyu Li, Zihui Song

Chunjiang Intelligence · Beijing Normal University

Efficient Long-Context Inference Peak Memory

May 29, 2026cs.CV

Light Interaction: Training-Free Inference Acceleration for Interactive Video World Models

Interactive video world models generate video chunk by chunk in response to user-controlled camera movements, enabling applications such as real-time game simulation, virtual scene navigation, and embodied AI training. However, scaling to long interactive trajectories is prohibitively expensive due to growing context memory, quadratic attention complexity, and repeated denoising steps. We present Light Interaction, a training-free inference acceleration framework for interactive video world models. Our key insight is that interaction naturally enables trajectory-dependent adaptive computation: retrieved spatial memory can be discarded during novel exploration, temporal context can be adjusted according to local latent dynamics, and early-step model outputs can be reused when the camera revisits familiar regions. Based on this insight, Light Interaction combines adaptive context management, denoising cache acceleration, and hardware-software co-designed 3D block sparse attention with fused Triton kernels. Evaluated on HY-WorldPlay and Matrix-Game-3.0, Light Interaction achieves up to 2.59x speedup without model retraining while maintaining competitive visual quality.

Jiacheng Lu, Haoyi Zhu, Sipei Yi +3

1Zhejiang University · 2NVIDIA

Video World Models Dynamic Sparse Attention Peak Memory

May 29, 2026cs.CV

Task-Focused Memorization for Multimodal Agents

Long-term memory is essential for multimodal agents to build coherent experience, accumulate world knowledge, and achieve continual learning. However, constructing effective memory goes beyond memory module design and basic requirements such as accuracy and fidelity; the key challenge lies in determining what to memorize. Multimodal agents, such as embodied agents, continuously perceive, reason, and act in real or virtual environments, receiving an unbounded stream of multimodal observations. From this combinatorial explosion of information, an agent must selectively retain content that is relevant to its role in the environment and valuable for future tasks. To bridge this gap, we frame memory generation as a learnable memorization policy and introduce TaskMem (Task-focused Memorization Policy Learning), a reinforcement-learning-based framework that enables the policy to dynamically adjust its focus to the demands of real tasks encountered in the environment. TaskMem adopts a two-phase training paradigm: Phase One learns how to memorize by optimizing memory quality under fundamental fidelity requirements; Phase Two occurs after deployment, where the agent learns what to memorize by tuning an adapter on its base MLLM, using recent environment tasks to define a reward model that guides the memorization policy toward task-relevant content. To evaluate our approach, we reformulate VideoMME, EgoLife, and EgoTempo into streaming benchmarks that simulate a realistic setting in which an agent processes streaming observations and handles tasks arriving online. To isolate memory assessment, the questions must be answered using only the agent's memory, without access to raw video. Built on Qwen3-VL-30B-A3B, TaskMem improves VQA accuracy by 6.3%, 7.0%, and 5.3% on these benchmarks, respectively.

Tao Zou, Yichen He, Tian Qiu +2

ByteDance Seed · Fudan University

Multimodal Memory Multimodal Agents Memorization Peak Memory Multi-Stage Training

May 29, 2026cs.CV

SlotMemory: Object-Centric KV Memory for Streaming Long-Video Generation

Streaming video generation models typically rely on temporal-centric memory, which organizes historical context as raw frames, chunk segments, or unclustered tokens. This organization frequently leads to identity drift and semantic inconsistency when entities exit the frame or during interactive prompt transitions. To address these limitations, we propose SlotMemory, an object-centric Key-Value memory mechanism for streaming video diffusion. Our approach shifts the memory abstraction from "when" an event occurred to "what" is being represented by decomposing the transformer's key-value manifold into discrete, reusable semantic slots. By utilizing these slots as routing addresses to index and store high-fidelity key-value tokens, we enable entity-level persistence and prompt-aware retrieval across long horizons. Evaluated on 60-second interactive narratives using the Wan2.1-T2V-1.3B backbone, SlotMemory achieves a state-of-the-art quality score of 81.61 and a 22.8 percent relative improvement in dynamic consistency over the strongest existing streaming baseline. Our results demonstrate that structured semantic representation, rather than raw temporal capacity, is the essential primitive for persistent long-form video synthesis. Our codes and checkpoints are available at https://tj12323.github.io/SlotMemory/.

Weijia Dou, Hui Li, Jiahao Cui +3

Fudan University · Meta Superintelligence Labs · Baidu

Interactive Video Generation Streaming Video Understanding Peak Memory

May 28, 2026cs.AR

Memory-Bound but Not Bandwidth-Limited: The Physical AI Inference Gap in Batch-1 LLM Decode

Physical AI systems, including robots, autonomous vehicles, embodied agents and edge copilots, often run a different inference workload from cloud LLM serving: single-stream, batch-1 autoregressive decode, where one robot, camera feed or user session waits on the next token. This workload is usually described as memory-bandwidth-bound. Each decode step streams model weights and the active KV cache, so latency should scale with peak HBM bandwidth. We show that this account is true but incomplete. We measure batch-1 decode for three 7 to 8B-class GQA transformers across four NVIDIA GPUs: H100 SXM5, A100-80GB SXM4, L40S and L4. We evaluate context lengths from 2048 to 16384, producing 44 valid cells under a controlled bf16 SDPA setup. The achieved fraction of peak HBM bandwidth falls as peak bandwidth rises. On the headline Qwen-2.5-7B ctx=2048 cell, an L4 reaches roughly 81 percent of its analytic memory floor, while an H100 reaches only 27 percent. Physical-AI decode is memory-dominated, but faster memory does not translate into proportional latency gains. We test the missing term with a CUDA Graphs A/B experiment. On H100 at ctx=2048, CUDA Graphs improves decode latency by 1.259x across N=10 fresh sessions, with a 95 percent bootstrap confidence interval of 1.253 to 1.267. On L4, the same intervention gives only 1.028x. This isolates a launch-side overhead that becomes visible on fast GPUs but remains mostly hidden on slower, bandwidth-bound GPUs. The deployment implication is that memory savings matter only when the runtime realises them. On L4, bf16 decode sits close to the memory floor, but common quantised paths do not recover the expected 4x weight-traffic reduction: bnb-nf4 reaches 59.36 ms/step and AutoAWQ+Marlin reaches 45.24 ms/step from a 62.32 ms bf16 baseline. GPTQ+ExLlamaV2, with Ada-tuned int4 kernels, reaches 17.36 ms/step.

Josef Chen

KAIKAKU

LLM Inference Optimization Graphics Processing Unit Memory Large Language Model Decoding Agentic Inference Peak Memory Semantic Memory

May 28, 2026cs.CV

OmniMem: Scalable and Adaptive Memory Retrieval for Long Video Generation

Autoregressive (AR) video generation extends videos by producing latent chunks sequentially, but scaling to long videos requires repeated access to a growing historical KV cache. Existing methods reduce this cost by truncating the KV cache or compressing it into implicit memory, but both lose explicit access to query-relevant historical details. We propose OmniMem, an explicit full-range memory retrieval framework that performs sparse KV retrieval over the historical cache. To make this practical for chunk-based AR video generation, OmniMem addresses two issues: (i) local bias in sparse KV selection and (ii) Union Explosion in memory access. Adaptive Window Exclusion removes local-window blocks from the selection candidates when sufficient long-range history is available, preserving the sparse budget for informative long-range retrieval. Query-Shared KV Selection reduces cross-query diversity, while Per-Head Scattered KV Access avoids expanding head-specific selections into a large selected KV buffer. This allows each attention head to retrieve non-contiguous KV blocks according to its own selection pattern. Experiments on long-video generation show that OmniMem improves Dynamic Degree by 52.3% and preserves strong consistency over strong baselines, while maintaining comparable memory usage.

Lin Zhao, Yushu Wu, Yifan Gong +2

Northeastern University · Adobe Research

Autoregressive Video Generation Long Video Generation Streaming Video Understanding Memory-Augmented Framework Video Generation Peak Memory+2

May 28, 2026cs.CL

HTAM: Hierarchical Transition-Attended Memory for Operator Optimization

High-performance GPU kernels are essential for efficient LLM deployment, yet optimizing them remains expertise-intensive. Recent LLM-based code generation makes automatic GPU operator generation promising, but operator optimization remains a hardware-aware search problem. Existing LLM-based methods face a granularity mismatch: coarse hints are reusable but hard to execute, whereas detailed memories are actionable but enlarge the search space and obscure optimization bottlenecks. The key challenge is therefore to organize optimization experience at an appropriate granularity. To address this issue, this paper proposes HTAM (Hierarchical Transition-Attended Memory), a coarse-to-fine framework for LLM-based operator optimization. HTAM builds a two-level Hierarchical Transition Graph (HTG) to organize coarse global directions, detailed local strategies, and transition experience between optimization steps. During each evolution step, HTAM selects a global direction from the current state and recent optimization history, retrieves the corresponding local strategy memory, and uses it to guide concrete CUDA code generation. Experiments on the full KernelBench suite demonstrate that HTAM consistently improves correctness, fast-solution rate, and speedup over LLM-based baselines, while backend and Robust-KBench studies indicate transferable benefits from structured memory.

Yining Zhang, Mingyang Yi, Chen Wang +5

School of Artificial Intelligence, University of Chinese Academy of Sciences · Institute of Automation, Chinese Academy of Sciences · Zhongguancun Academy +1

Code Optimization Peak Memory Bottlenecks Hierarchical

May 28, 2026cs.AI

VikingMem: A Memory Base Management System for Stateful LLM-based Applications

Large Language Models have revolutionized interactive applications; however, their finite context windows pose a critical data management challenge for maintaining stateful, long-term interactions. Existing memory approaches often rely on simplistic extraction methods that lead to incomplete memories or use rigid, single-purpose memory extraction prompts tailored to a single use case, such as chatbots. Consequently, they lack generalizability and perform poorly across diverse downstream tasks. To bridge this gap, we introduce the Memory Base, a novel data management paradigm for managing the persistent state of long-term interactions. It is characterized by three core principles: selective extraction of high-value memories from raw information streams; inherent statefulness and evolution, where memory content is progressively summarized, corrected, and temporally weighted to prioritize recent interactions; and a generalizable abstraction paradigm designed for robust transferability across diverse applications, including education, recommendation, and agent memory. Building on this foundation, we present VikingMem, an end-to-end Memory Base Management System implemented on the VikingDB vector engine. VikingMem materializes this paradigm through interconnected event and entity abstractions. It features event-centric memory extraction to selectively handle complex information streams, while entities are dynamically updated by events to achieve stateful evolution. Using temporal compression via a topic-wise timeline and time-weighted recall, the system progressively produces high-level summary memories, prioritizes recent items, and compresses and fades older ones. Extensive evaluations on long-term memory benchmarks demonstrate that VikingMem outperformes baselines by up to 30% in memory retrieval effectiveness while maintaining the low latency essential for interactive applications.

Jiajie Fu, Junwen Chen, Mengzhao Wang +5

Zhejiang University · China · ByteDance

May 28, 2026cs.CL

Learning Design Skills as Memory Policies for Agentic Photonic Inverse Design

Photonic crystal fiber (PCF) inverse design remains challenging because candidate geometries must satisfy coupled optical targets under expensive electromagnetic simulation. Existing pipelines improve surrogate prediction or one-shot parameter recommendation, but they do not accumulate reusable design knowledge across iterative trials. We formulate PCF inverse design as a memory-policy learning problem and propose SkillPCF, a closed-loop agent framework that combines a physics-guided memory skill bank, reinforcement-learned skill selection, and simulator-grounded skill evolution. We further construct a real-world dataset with 479 expert interaction traces (2,507 spans) and 553 memory-dependent evaluation queries covering dispersion engineering, loss optimization, and multi-objective design. Experiments across multiple LLM backbones and classical baselines show that SkillPCF achieves stronger design-quality and efficiency trade-offs under practical simulation budgets, demonstrating the effectiveness of our proposed memory-skill learning paradigm for physics-aware PCF inverse design.

Shengchao Chen, Ting Shu, Sufen Ren

AAII, University of Technology Sydney · School of Artificial Intelligence, Shenzhen University · School of Information and Communication Engineering, Hainan University

Nanophotonic Inverse Design Skills Peak Memory Large Language Model Backbones

May 27, 2026cs.LG

LoRe: Adaptive Interaction-Evaluation Routing with Per-Step Interaction Budgets for Iterative Graph Solvers

Diffusion-based neural solvers for combinatorial optimization repeatedly re-evaluate dense edge/factor interactions, making inference expensive in wall-clock time and often memory-bound at scale. Inspired by the computational methodologies of many-body physics, we introduce LoRe, a training-free, inference-time drop-in wrapper that enforces per-step interaction-evaluation budgeting: at each iteration, it evaluates only a fixed fraction of interactions by dynamically routing computation to high-conflict or high-uncertainty interactions, instead of using a fixed sparsification (e.g., static kNN graphs or static masks). Under fully inclusive end-to-end wall-clock accounting, LoRe substantially improves scalability on the Maximum Independent Set (MIS) problem, extending feasible inference more than

3\times

beyond the baseline's out-of-memory limit, delivering a

\sim 8\times

speedup and a

\sim 12\times

peak-memory reduction, with solution quality preserved in this regime. Demonstrating cross-task generality on the large-scale Traveling Salesperson Problem (TSP) and zero-shot robustness to topology shifts, LoRe achieves a

\sim 15\times

speedup at

n=1000

with a

44\times

memory reduction and competitive tour quality.

Jintao Li, Yong-Yi Wang, Zheng-An Wang +1

Beijing Key Laboratory of Fault-Tolerant Quantum Computing, Beijing Academy of Quantum Information Sciences, Beijing, China · Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, CAS, Beijing, China · Beijing Key Laboratory of Advanced Quantum Technology, Beijing, China +1

Neural Solvers Solver Iterations Inference-Time Token Budget Allocation Peak Memory Many-Body Physics

May 27, 2026cs.LG

Augmenting Attention with Exponentially Decaying Memory Improves Query-Aware KV Sparsity

Efficient inference is critical for long-context language models, where attention computation and KV-cache access dominate the cost. Recent work RAT+, introduces a recurrence-augmented attention backbone that enables flexible dilated attention at inference time. In this paper, we investigate whether this exponentially decaying memory can also improve existing query-aware sparse inference methods. Using representative methods including Quest, MoBA, and SnapKV, we show that RAT+ consistently improves accuracy over standard attention across sparse budgets on eight needle-in-a-haystack tasks. We validate these gains both on the released checkpoints from the RAT+ paper and on OLMo2-7B, which we continue pretraining with the added memory module for 10B tokens. Finally, we propose two hypotheses explaining why this memory module benefits query-aware sparse inference and design targeted experiments to support them.

Xiuying Wei, Caglar Gulcehre

EPFL, CLAIRE Lausanne, Switzerland

Dynamic Sparse Attention Efficient Long-Context Inference Peak Memory Sparsity Inference-Time

May 27, 2026cs.LG

Outer-Momentum Restarting in High-Dimensional Two-Phase Optimization

Communication-efficient distributed optimizers such as DiLoCo reduce synchronization costs by letting workers perform many local updates before aggregating their progress with an outer momentum optimizer. Recent theory suggests that the outer optimizer acts on an effective spectrum induced by the inner optimization loop, and that the choice of outer momentum controls how progress from local updates is accumulated across communication rounds. We study periodic restarting of the outer momentum as a simple complementary mechanism for controlling this outer memory. In a linearized squared-loss model where prediction-space residuals evolve under the empirical NTK, we derive a mode-wise restart contraction showing that resets exploit phase cancellation by discarding stale momentum while preserving inner-loop progress. Toy experiments verify the predicted contraction behavior, and language-model pretraining experiments show that periodic restarts widen the stable range of outer learning rates and momentum values across communication periods.

Kristi Topollai, Allan Ma, Tolga Dimlioglu +2

New York University, New York, USA

Momentum Peak Memory

May 27, 2026cs.AI

GONDOR to the Rescue: Satisficing Planning with Low Memory

Greedy Best-First Search (GBFS) is the dominant approach for solving search problems where the goal can be estimated with a heuristic, such as planning, route finding, navigation, and pathfinding. This is especially true when the memory is tightly constrained, such as planning on edge devices. To alleviate that, we present GONDOR (Greedy Online Navigation with Dynamic Outpost-based Re-search), a memory-efficient extension of GBFS that allows search to continue under strict memory limits by periodically compressing the search tree while retaining a sparse set of anchor states, then upon reaching the goal reconstructs the path by re-searching between the sparse states. We analyze the algorithm and discuss several variants defined by different outpost selection policies. In addition, we explore using Bloom filters for compact duplicate detection in the closed list. Experiments across numeric planning domains and heuristic configurations show that GONDOR consistently improves coverage under low memory budgets compared to standard GBFS. We release the implementation of GONDOR and the Bloom-filter variant to facilitate further research on memory-efficient heuristic search.

Yonatan Vernik, Alexander Tuisov, Alexander Shleyfman

Computer Science Department, Bar-Ilan University · Independent Researcher

Path Planning Planning Peak Memory

May 27, 2026cs.LG

Context Distillation as Latent Memory Management

Context distillation compresses contextual information into model parameters, yet existing methods often ignore how multiple distilled latent memories should be stored, retrieved, and safely activated in non-oracle settings. We formulate context distillation as a latent memory management problem. We distill each context into an independent LoRA adapter, forming a modular memory bank that enables explicit memory selection. Given a query, our framework retrieves candidate memories, routes the query to the most suitable adapter, and uses a Self-Gating mechanism to decide whether latent memory should be activated. To improve efficiency, we further introduce cache sharing to reduce management overhead during inference. Experiments show that our method substantially outperforms baselines with retrieval, while Self-Gating improves robustness by deactivate unnecessary latent memories.

Ziyang Zheng, Zeju Li, Xiangyu Wen +5

The Chinese University of Hong Kong · Huawei Noah’s Ark Lab

Dataset Distillation Peak Memory Low-Rank Adaptation Adapters

May 27, 2026cs.CL

ConvMemory: A Lightweight Learned Memory Reranker, a Negative Attribution Result, and a Research-Preview Conflict Editor

We describe ConvMemory, a small 3.6M-parameter learned reranker for conversational long-term memory retrieval, trained with cross-encoder teacher supervision over fused dense and lexical features. On the LongMemEval memory family, ConvMemory operates above the BGE-large cross-encoder in Recall@10 at 12-47x lower latency, remains within 0.025 Recall@10 of mxbai-rerank-large-v1 on Clean500 while running 28x cheaper; under Stress1000 distractors the Recall@10 gap widens to 0.081 but ConvMemory still operates at 117x lower latency; these LongMemEval numbers are single-run or single-seed and are reported as indicative cost-frontier evidence, not benchmark-grade. We then publish a rigorous negative attribution result on a previously claimed mechanism: a five-seed retrained ablation with paired bootstrap shows that ConvMemory's learned temporal window is statistically significant on aggregate but not temporally specific, with the largest effects on hard non-temporal controls and no significant effect on multi-hop temporal queries. The honest description of the mechanism is cheap cross-encoder distillation in a fused dense+lexical feature space, not temporal-structure exploitation. We additionally release CCGE-LA, a low-amplitude conflict-aware candidate-set editor over ConvMemory, as a research preview with modest but consistent gains on supersession and stale/rescue slices on LoCoMo. All results are retrieval-stage; ConvMemory does not match mxbai-rerank-large-v1 in absolute LoCoMo MRR, and the report is single-author and not yet independently audited.

Taiheng Pan

School of Computing and Information Systems University of Melbourne

Conversational Memory Cross-Encoder Peak Memory Cross-Encoder Reranking Reranking Negative Results+2

May 27, 2026cs.AI

MemCog: From Memory-as-Tool to Memory-as-Cognition in Conversational Agents

Existing agent memory systems universally follow what we term a Memory-as-Tool paradigm where a single query triggers one-shot retrieval of flat passage lists, suffering from passive invocation, reasoning-retrieval decoupling, and structural mismatch between retrieved fragments and the agent's navigational needs. We propose MemCog, a Memory-as-Cognition system that makes memory access an integral part of the reasoning process. MemCog organizes user knowledge as Navigable Memory Store with associative link graphs, exposes Cross-Dimensional Navigation Interface for multi-step reasoning-driven traversal, and employs Proactive Reasoning Protocol that drives agents to spontaneously initiate memory exploration from conversational context. We additionally construct ProactiveMemBench, the first benchmark for evaluating proactive memory triggering. Experiments show that MemCog achieves state-of-the-art on passive QA benchmarks (92.98 on LoCoMo, 95.8 on LongMemEval) while substantially outperforming baselines on ProactiveMemBench, demonstrating the advantage of Memory-as-Cognition.

Zihan Li, Xingyu Fan, Feifei Li +1

WeChat, Tencent Inc., Beijing, China

Cogadapt Peak Memory Traversal

May 27, 2026cs.CL

MemGuard: Preventing Memory Contamination in Long-Term Memory-Augmented Large Language Models

Memory-augmented large language models extend reasoning beyond a fixed context window by maintaining long-term memory across interactions. However, existing memory systems often collapse stable user facts, episodic events, and behavioral rules into a shared space, allowing functionally distinct memories to be retrieved and used as interchangeable evidence. We identify this failure mode as heterogeneous memory contamination, where context-specific events become overgeneralized claims, or semantically relevant but functionally incompatible memories mislead generation. To this end, we introduce MemGuard, a type-aware memory framework that preserves functional memory boundaries during memory construction and retrieval. It assigns each memory an explicit functional role at write time, maintains relations across type-isolated memories, and selectively composes evidence only from necessary memory types, reducing contamination from irrelevant or functionally incompatible evidence. Across hallucination and long-horizon conversation benchmarks, MemGuard improves memory reliability by up to 28.27% while retrieving up to 5.8x fewer memory tokens than prior methods. These results suggest that reliable long-term reasoning depends on principled organization and selective use of heterogeneous memory.

Hyeonjeong Ha, Jeonghwan Kim, Cheng Qian +7

University of Illinois Urbana-Champaign · Capital One · Columbia University

Long-Term Memory Peak Memory Long-Term

May 25, 2026cs.CL

Memory Architectures for Multi-Turn Text-to-SQL: A Benchmark and Empirical Study

Multi-turn Text-to-SQL is central to enterprise analytics yet remains predominantly evaluated in single-turn settings. We introduce EnterpriseMem-Bench, a multi-turn Text-to-SQL benchmark of 300 sessions and 1,400 turns built programmatically from three enterprise domains (BIRD financial, SEC EDGAR, Northwind), with deterministic ground truth and per-turn memory-critical annotation. We evaluate five frontier models -- GPT-5 mini, GPT-5.2, Claude Sonnet 4.5, Sonnet 4.6, and Opus 4.6 -- across five memory conditions enabling a three-way ablation isolating working-memory window size, episodic retrieval, and semantic augmentation as independent effects. All Claude models are evaluated with extended thinking enabled to maintain parity with GPT reasoning models. We introduce the Memory Benefit Score (MBS) as a per-turn diagnostic metric. Four findings emerge: (1) stateless multi-turn Text-to-SQL collapses to zero execution accuracy by Turn 3 across all five models, even under reasoning; (2) memory-architecture complexity does not monotonically improve accuracy -- working memory dominates, and additional components produce model- and dataset-dependent effects from +14 to -16 percentage points; (3) Claude Sonnet 4.6 underperforms Sonnet 4.5 by 17-33pp on SEC EDGAR across conditions, a generational regression persisting under reasoning; (4) under reasoning, Claude error distributions become mono-modal -- every non-correct turn is a wrong-result error. We release the benchmark, agent, and evaluation code.

Ravi Kumar Tummalapenta, Suman Addanki

LLM Suite Engineering Team, JP Morgan Chase & Co.

Text-To-Sql Large Language Model Memory Peak Memory Empirical Study Analytics

May 25, 2026cs.AI

Personalize-then-Store: Benchmarking and Learning Personalized Memory for Long-horizon Agents

Existing large language model (LLM) based memory systems apply universal, static policies that overlook a fundamental reality: the contexts that are worth storing in memory are different across users. This misalignment wastes limited memory budget on transient interactions while failing to preserve critical context for long horizon tasks. To address this gap, we investigate an underexplored question: can LLM based memory systems learn personalized memory policies? We introduce PerMemBench, the first benchmark for evaluating personalized memory systems, featuring multi year, multi domain interaction histories across diverse user personas. We further present the first empirical study of memory personalization, proposing session level storage gating, a lightweight framework that selectively bypasses memory operations for transient sessions. Our study confirms that personalization yields substantial retention gains under perfect gating, yet reveals that accurate gating remains an open and critical challenge.

Yeonjun In, Wonjoong Kim, Sangwu Park +2

KAIST

Interaction History Peak Memory Personalization Long-Horizon Agents

May 21, 2026cs.LG

Asymmetric Virtual Memory Paging for Hybrid Mamba-Transformer Inference

Hybrid language models like Jamba mix attention layers with State Space Models (SSMs), creating two memory cache types with opposite profiles: Key-Value (KV) caches grow linearly with sequence length, while SSM states stay fixed per layer. Current inference engines handle this poorly. Unified pools pad SSM states to attention page sizes, wasting up to 7.3x capacity. Static dual pools cannot adapt when prompt distributions shift between requests. We present Asymmetric Virtual Memory Paging (AVMP). The allocator separates the two cache types into physically distinct pools behind a unified virtual address space, and migrates capacity between pools when one runs out. Migration triggers only on allocation failure, keeping behavior deterministic. We evaluate AVMP across 270 synthetic cells plus 60 cells of ShareGPT trace replay on an RTX 3060 12GB. Out-of-Memory events drop 7.6% and request throughput improves 1.83x to 13.3x across synthetic workloads and 2.36x on ShareGPT. All gains hold under paired-bootstrap 95% confidence intervals. A phase-time breakdown reveals two distinct mechanisms: shorter OOM recovery on capacity-pressured workloads, and faster allocation calls on KV-heavy workloads. Implementation is pure Python; Triton integration is future work.

An Xuan Nguyen

Codepawl Ho Chi Minh City, Vietnam

Hybrid Mamba-Transformer Model Peak Memory Duality

May 21, 2026cs.LG

Short-Term-to-Long-Term Memory Transfer for Knowledge Graphs under Partial Observability

Reinforcement learning under partial observability requires deciding what information to retain, yet most memory-based approaches do not explicitly model short-term-to-long-term transfer of symbolic observations. We study this transfer process in a temporal knowledge-graph memory setting and cast it as a neuro-symbolic value-based decision problem: for each observed triple, the agent chooses whether to keep or drop it before long-term insertion. To handle variable-sized short-term buffers, we use a per-item Q-learning design with shared parameters and a practical temporal-difference update over matched items across consecutive steps. On the RoomKG benchmark at long-term memory capacity 128, learned transfer decisions outperform symbolic and neural baselines, including symbolic baselines with temporal annotations and history-based LSTM/Transformer baselines. Across transfer-policy ablations, a lightweight local short-term-only variant performs best, and step-level behavior shows that the policy keeps navigation- and query-relevant facts while discarding lower-value candidate facts, supporting explicit and interpretable memory decisions under memory constraints.

Taewoon Kim, Vincent François-Lavet, Michael Cochez

1HumemAI · 2Vrije Universiteit Amsterdam · 3ELLIS Institute Finland & Abo Akademi University

Temporal Knowledge Graphs Long-Term Memory Knowledge Graphs Peak Memory Partial Observability Neuro-Symbolic Framework

May 19, 2026cs.LG

SMA-DP: Spectral Memory-Aware Differential Privacy for Deep Learning

Differentially private stochastic gradient descent (DP-SGD) enables private deep learning through per-example clipping and calibrated Gaussian noise, but its high-variance updates can reduce utility on challenging datasets. We propose \textbf{SMA-DP-SGD}, a \textbf{Spectral Memory-Aware Differentially Private Stochastic Gradient Descent} method that augments DP-SGD with a fractional memory branch built only from previously privatized noisy releases. WeightWatcher-inspired power-law spectral exponents provide group-wise reliability signals, instantiated layer-wise in our experiments, to adapt the decay and effective memory depth. Private-history alignment, norm matching, and warm-up activation stabilize the memory contribution. Privacy remains transparent: conditioned on the private release history, the memory branch is fixed, and the only newly data-dependent term is the current clipped sum scaled by a fixed coefficient

β

. Hence, SMA-DP-SGD preserves a clean conditional sensitivity structure and exactly recovers group-wise DP-SGD when

β=1

. Experiments on CIFAR-100, CIFAR-10, and MNIST show competitive or superior accuracy over several DP optimization baselines, with the largest gains on CIFAR-100 and CIFAR-10. CIFAR-10 ablations show that

β

controls the privacy--utility trajectory, while spectral and memory diagnostics confirm a controlled short-to-moderate effective memory depth and a small memory-branch ratio. Runtime analysis shows that the mechanism incurs additional overhead, about

2.94\times

DP-SGD in our CIFAR-10 implementation, revealing a practical trade-off between adaptive private memory and computational cost.

Mohammad Partohaghighi, Roummel Marcia

Department of Electrical Engineering and Computer Science University of California, Merced Merced, CA, USA · Department of Applied Mathematics University of California, Merced Merced, CA, USA

Stochastic Gradient Descent Privacy Deep Learning Peak Memory

May 19, 2026cs.CV

GoTTA be Diverse: Rethinking Memory Policies for Test-Time Adaptation

Test-time adaptation (TTA) enables a pre-trained model to adapt online to an unlabeled test stream under distribution shift. While most TTA research focuses on the adaptation objective, practical streams also depend critically on the memory used to select which test samples drive adaptation. Existing memory mechanisms are usually evaluated as components of specific TTA algorithms, making it difficult to isolate which memory design choices matter and when they matter. In this work, we provide a systematic benchmark that decouples memory from the adaptation algorithm and evaluates memory policies under unified conditions across i.i.d., non-i.i.d., continual, and practical test streams. Our study shows that effective memory management requires more than retaining recent or class-balanced samples. In particular, intra-class diversity is a key factor for avoiding redundant buffers and maintaining representative adaptation signals under temporally correlated and label-skewed streams. Motivated by this finding, we introduce Guided Observational Test-Time Adaptation (GOTTA), a family of diversity-aware memory policies that combine class-balanced allocation with feature-space diversity. GOTTA memories act as drop-in replacements for existing buffers and can be paired with different TTA objectives. Across corruption benchmarks and video-stream settings, diversity-aware memory improves adaptation most clearly under constrained memory budgets and challenging non-i.i.d. streams, while remaining competitive as memory capacity increases. These results highlight memory management as a first-class component of robust test-time adaptation and identify diversity as a central principle for practical TTA.

Shyma Alhuwaider, Yasmeen Alsaedy, Merey Ramazanova +2

Center of Excellence in Generative AI, KAUST, Saudi Arabia

Test-Time Adaptation Peak Memory Unlabeled Data

May 18, 2026cs.LG

KVBuffer: IO-aware Serving for Linear Attention

Linear attention has recently gained significant attention for long-context inference due to its constant decoding cost with respect to context length. However, existing serving systems typically serve linear attention by recurrently computing and updating a large linear attention state in every decoding step. Since the state is much larger than the per-token key and value, recurrent decoding incurs substantial memory access and becomes inefficient for serving linear attention. In this paper, we propose KVBuffer, an IO-aware serving mechanism for linear attention. By buffering recent keys and values, KVBuffer enables serving systems to compute linear attention outputs in more flexible and memory-efficient ways. For decoding, KVBuffer enables chunkwise computation, which reduces average memory access and decoding latency by deferring state updates and applying them in batch. For speculative decoding, KVBuffer verifies draft tokens in parallel and avoids storing temporary states. For short contexts, KVBuffer computes attention outputs directly from buffered keys and values, without creating or updating the linear attention state. We implement KVBuffer in SGLang for Qwen3-Next. Our evaluations show that KVBuffer can reduce linear attention decoding latency by up to 45.17% and increase the maximum number of serving requests by 5x for speculative decoding when verifying four draft tokens.

Longwei Zou, Lin Zhong

Department of Computer Science Yale University New Haven, CT 06511

Kimi Delta Attention Depthweave-Kv Large Language Model Serving Linear Attention Speculative Decoding Peak Memory

May 18, 2026cs.CV

MementoGUI: Learning Agentic Multimodal Memory Control for Long-Horizon GUI Agents

Recent GUI agents have made substantial progress in visual grounding and action prediction, yet they remain brittle in long-horizon tasks that require maintaining task state across many interface transitions. Existing agents typically rely on raw history replay or text-only memory, which either overwhelms the model with redundant screenshots or discards localized visual evidence needed for future decisions. To address these limitations, we introduce \textbf{MementoGUI}, a plug-in agentic memory framework that equips MLLM-based GUI agents with \textbf{MementoCore}, a learned controller for online memory selection, compression, and retrieval. Rather than treating interaction history as a fixed context, MementoGUI formulates long-horizon GUI control as an online memory-control problem: working memory selectively preserves task-relevant interface events with textual summaries and ROI-level visual evidence, while episodic memory retrieves reusable past trajectories through learned relevance selection. MementoCore modularizes memory control into specialized operators for step processing, memory compression, episodic writing, and episodic selection, enabling plug-in memory augmentation without finetuning the GUI agent backbone. We further develop a scalable data curation pipeline that converts computer-use trajectories into memory-controller training data, introduce \textbf{MementoGUI-Bench} for evaluating long-horizon decision-making in GUI agents, and design MLLM-based metrics for semantic action matching, task progress, and memory consistency. Experiments on GUI-Odyssey, MM-Mind2Web, and MementoGUI-Bench show that MementoGUI consistently improves GUI agents over no-history, history-replay, and text-only memory baselines, with larger MementoCore backbones further strengthening memory-augmented GUI control.

Ziyun Zeng, Hang Hua, Bocheng Zou +3

University of Rochester · MIT-IBM Watson AI Lab · University of Wisconsin-Madison

Agentic Memory Graphical User Interface Episodic Memory Peak Memory Action Prediction

May 18, 2026cs.CV

TinySAM 2: Extreme Memory Compression for Efficient Track Anything Model

Segment Anything Model 2 (SAM 2) serves as a core foundation model in the field of video segmentation. Building upon the original SAM model, it introduces a memory bank mechanism and demonstrates outstanding performance in tasks such as semi-supervised video object segmentation and tracking anything. However, the complex computational characteristics of SAM 2's multi-stage image encoder and memory module have raised the barrier to the model's deployment in practical applications. To address this issue, we propose TinySAM 2, a lightweight video segmentation model that balances performance and efficiency. First, a memory quality management mechanism is introduced to select and retain high-informative historical frames as the memory. In addition, a joint-spatial-temporal token compression is proposed that reduces the memory storage and computational cost. Specifically, average pooling is employed to first compress redundancy tokens in the spatial domain. In the temporal domain, informative tokens are selected across frames in the memory bank based on token-level similarity measurement. Besides, we take RepViT as the lightweight image encoder, which further reduces the model parameters. Extensive experiments on challenging datasets such as DAVIS and SA-V demonstrate that TinySAM 2 achieves 90% of the performance of SAM 2.1, with only 7% memory tokens and 3% training data. This study effectively alleviates the bottlenecks in parameter count, computational load, and deployment costs associated with SAM 2, providing a resource-efficient solution for the widespread application of video segmentation models on devices.

Zhaoyuan Ding, Yijing Yang, Han Shu +1

Huawei Noahs Ark Lab

Video Object Segmentation Segment Anything Model Token Compression Video Compression Data Compression Methods Peak Memory

May 13, 2026cs.LG

MinT: Managed Infrastructure for Training and Serving Millions of LLMs

We present MindLab Toolkit (MinT), a managed infrastructure system for Low-Rank Adaptation (LoRA) post-training and online serving. MinT targets a setting where many trained policies are produced over a small number of expensive base-model deployments. Instead of materializing each policy as a merged full checkpoint, MinT keeps the base model resident and moves exported LoRA adapter revisions through rollout, update, export, evaluation, serving, and rollback, hiding distributed training, serving, scheduling, and data movement behind a service interface. MinT scales this path along three axes. Scale Up extends LoRA RL to frontier-scale dense and MoE architectures, including MLA and DSA attention paths, with training and serving validated beyond 1T total parameters. Scale Down moves only the exported LoRA adapter, which can be under 1% of base-model size in rank-1 settings; adapter-only handoff reduces the measured step by 18.3x on a 4B dense model and 2.85x on a 30B MoE, while concurrent multi-policy GRPO shortens wall time by 1.77x and 1.45x without raising peak memory. Scale Out separates durable policy addressability from CPU/GPU working sets: a tensor-parallel deployment supports 10^6-scale addressable catalogs (measured single-engine sweeps through 100K) and thousand-adapter active waves at cluster scale, with cold loading treated as scheduled service work and packed MoE LoRA tensors improving live engine loading by 8.5-8.7x. MinT thus manages million-scale LoRA policy catalogs while training and serving selected adapter revisions over shared 1T-class base models.

Mind Lab, :, Song Cao +60

May 13, 2026cs.NE

WASHH: An Anchor-Aware Whale-Guided Selection Hyper-Heuristic for Continuous Optimization and SVC Configuration

Learning-assisted algorithm design often has to make reliable search decisions under small evaluation budgets, where committing to a single metaheuristic can be unreliable. We propose WASHH, a Whale-guided Adaptive Selection Hyper-Heuristic for continuous black-box optimization. WASHH uses WOA as the main exploitation backbone, but treats PSO-style memory, GWO-style leader averaging, DE-style variation, local coordinate search, and anchor-guided refinement as selectable search behaviors. An online reward controller allocates evaluations according to observed improvements, while anchor refinement exploits inexpensive reference configurations such as box centers or default model settings without bypassing black-box evaluation. On ten 30-dimensional benchmark functions with 10 independent runs and 12,000 evaluations, WASHH achieves the best average rank, 1.10, and is best or tied best on all ten functions. It strictly improves over WOA on eight functions and ties WOA at the numerical optimum on Rastrigin and Griewank. We further study SVC hyperparameter configuration for breast cancer diagnosis under a 300-evaluation budget. WASHH obtains the lowest mean validation log loss among the compared optimizers, suggesting that anchor-aware selection hyper-heuristics are a practical lightweight direction for LEAD systems.

Yifu Zhao, Xiaofan Zou, Junhao Wei +9

Faculty of Applied Sciences, Macao Polytechnic University, Macao, 999078, China · School of Mechanical and Electrical Engineering and Automation, Shanghai University, Shanghai, 200444, China · Pazhou Lab (Huangpu), Guangzhou, 510555, China +2

Black-Box Optimization Metaheuristics Alignment-Aware Selection Heuristics Sparse Spatial Anchors Peak Memory

May 13, 2026cs.LG

Contextual Bandits for Resource-Constrained Devices using Probabilistic Learning

Contextual bandits (CB) are online sequential decision-making problems under partial feedback that underpin many adaptive services. There is a growing demand to deploy CB agents directly on-device, under strict constraints on memory, compute, and energy. However, standard linear CB algorithms are often impractical for resource-constrained devices with their unfavorable scaling in computational and memory costs. Recently, HD-CB, a CB approach based on hyperdimensional computing principles, has been proposed to model and solve CB problems by moving into high-dimensional spaces. HD-CB offers faster convergence, favorable scalability, and improves memory efficiency compared to linear CB algorithms. However, its learning rule is accumulation-based: the values of action vectors grow over time, requiring high precision. While periodic binarization can prevent overflow in low-precision components, it may discard important information about magnitudes and degrade decision quality. This paper introduces probabilistic HD-CB, a low-precision variant that replaces deterministic accumulation with a probabilistic update rule. At each step, only a random subset of vector components is updated, with a time-decaying update probability, and component values are constrained to a predefined range [-k,+k]. This approach enables low-precision components, prevents overflow without periodic binarization, and reduces the expected update cost in proportion to the fraction of updated components. Off-policy evaluation on standardized synthetic CB benchmarks using the Open Bandit Pipeline shows that probabilistic HD-CB consistently outperforms binarized HD-CB at equal precision, while approaching the performance of HD-CB with as few as 3 bits per component.

Marco Angioli, Kevin Johansson, Antonello Rosato +2

Sapienza University of Rome · Orebro University · Linkoping University +1

Contextual Bandit Framework Resource-Constrained Edge Devices Sequential Decision Making Probabilistic Model Peak Memory

May 13, 2026cs.LG

Continual Fine-Tuning of Large Language Models via Program Memory

Parameter-Efficient Fine-Tuning (PEFT), particularly Low-Rank Adaptation (LoRA), has become a standard approach for adapting Large Language Models (LLMs) under limited compute. However, in continual settings where models are updated sequentially with small datasets, conventional LoRA updates struggle to balance rapid adaptation and knowledge retention. Existing methods typically treat the low-rank space as a homogeneous update region, lacking mechanisms to regulate how short-term updates are consolidated over time. We propose a continual LoRA framework with \textbf{Pro}gram memory, inspired by \textbf{C}omplementary \textbf{L}earning Systems in neuroscience. Our approach, dubbed \textbf{ProCL}, organizes LoRA adapters into structured program memory slots that are dynamically retrieved through input-conditioned attention. This enables rapid and localized adaptation, encouraging similar inputs to reuse shared adapter regions while reserving unused capacity for future data. The slots are then combined with the underlying adapter, which maintains a distributed representation that gradually accumulates knowledge across tasks to balance plasticity and stability. Our method operates entirely within the LoRA parameterization and incurs no additional inference cost. Experiments on diverse benchmarks demonstrate improved retention and reduced catastrophic forgetting over other continual LoRA strategies.

Hung Le, Svetha Venkatesh

Deakin Applied AI Initiative, Deakin University, Australia

Training-Side Stage-Aware Low-Rank Adaptation Replay-Based Continual Learning Catastrophic Forgetting Model Fine-Tuning Continual Fine-Tuning Peak Memory+1

May 12, 2026quant-ph

Controllable Quantum Memory Capacity in Quantum Reservoir Networks with Tunable partial-SWAPs

In the field of quantum reservoir computing (QRC), many different computational models and architectures have been proposed. From these models, we identify feedback-based models -- which use a feedback mechanism to re-embed classical measurements from the QRC -- and recurrent models -- which use a multi-register approach with memory and readout qubits -- as the two major competing architectures that have been discussed and validated on hardware. In this paper, we advance upon the recurrent architectures, which employ a two register approach to endow the QRC with a fading memory. While these approaches have been validated on hardware and have demonstrated great real-world performance on noisy-intermediate-scale-quantum (NISQ) quantum processing units (QPUs), the exact mechanism through which the memory capacity arises is not completely understood or fully controllable. With this, we augment the recurrent approaches and present a hardware-realizable mechanism, which we call a tunable partial-SWAP, that allows for the direct control of the rate of memory dissipation from a QRN implemented on a gate-based QPU. The theory behind this mechanism is discussed in terms of a controlled amplitude-damping channel and validation experiments using a randomized short-term memory capacity (STMC) recall benchmark and the NARMA-5 dataset are conducted using simulation and IBM QPUs, respectively.

Erik L. Connerty, Ethan N. Evans

University of South Carolina - Columbia, Columbia, SC, USA · Qodex Quantum, Chicago, IL, USA

Reservoir Computing Peak Memory Switch

May 12, 2026cs.LG

MEME: Multi-entity & Evolving Memory Evaluation

LLM-based agents increasingly operate in persistent environments where they must store, update, and reason over information across many sessions. While prior benchmarks evaluate only single-entity updates, MEME defines six tasks spanning the full space defined by the multi-entity and evolving axes, including three not scored by prior work: Cascade and Absence (dependency reasoning) and Deletion (post-removal state). Evaluating six memory systems spanning three memory paradigms on 100 controlled episodes, we find that all systems collapse on dependency reasoning under the default configuration (Cascade: 3%, Absence: 1% in average accuracy) despite adequate static retrieval performance. Prompt optimization, deeper retrieval, reduced filler noise, and most stronger LLMs fail to close this gap. Only a file-based agent paired with Claude Opus 4.7 as its internal LLM partially closes the gap, but at ~70x the baseline cost, indicating closure currently depends on configurations that are not practical at scale. Code and data are available on the project page: https://seokwonjung-jay.github.io/meme-eval/.

Seokwon Jung, Alexander Rubinstein, Arnas Uselis +2

1KAIST AI · 2Tübingen AI Center, University of Tübingen · 3NAVER AI Lab

Peak Memory Memes Entities

May 12, 2026cs.AI

$δ$ -mem: Efficient Online Memory for Large Language Models

Large language models increasingly need to accumulate and reuse historical information in long-term assistants and agent systems. Simply expanding the context window is costly and often fails to ensure effective context utilization. We propose

δ

-mem, a lightweight memory mechanism that augments a frozen full-attention backbone with a compact online state of associative memory.

δ

-mem compresses past information into a fixed-size state matrix updated by delta-rule learning, and uses its readout to generate low-rank corrections to the backbone's attention computation during generation. With only an

8\times8

online memory state,

δ

-mem improves the average score to

1.10\times

that of the frozen backbone and

1.15\times

that of the strongest non-

δ

-mem memory baseline. It achieves larger gains on memory-heavy benchmarks, reaching

1.31\times

on MemoryAgentBench and

1.20\times

on LoCoMo, while largely preserving general capabilities. These results show that effective memory can be realized through a compact online state directly coupled with attention computation, without full fine-tuning, backbone replacement, or explicit context extension.

Jingdi Lei, Di Zhang, Junxian Li +7

Nanyang Technological University · Mind Lab · Fudan University +3

Peak Memory Efficient Long-Context Inference Kimi Delta Attention Frozen Backbone Memoryagentbench

May 12, 2026cs.CV

Robust Promptable Video Object Segmentation

The performance of promptable video object segmentation (PVOS) models substantially degrades under input corruptions, which prevents PVOS deployment in safety-critical domains. This paper offers the first comprehensive study on robust PVOS (RobustPVOS). We first construct a new, comprehensive benchmark with two real-world evaluation datasets of 351 video clips and more than 2,500 object masks under real-world adverse conditions. At the same time, we generate synthetic training data by applying diverse and temporally varying corruptions to existing VOS datasets. Moreover, we present a new RobustPVOS method, dubbed Memory-object-conditioned Gated-rank Adaptation (MoGA). The key to successfully performing RobustPVOS is two-fold: effectively handling object-specific degradation and ensuring temporal consistency in predictions. MoGA leverages object-specific representations maintained in memory across frames to condition the robustification process, which allows the model to handle each tracked object differently in a temporally consistent way. Extensive experiments on our benchmark validate MoGA's efficacy, showing consistent and significant improvements across diverse corruption types on both synthetic and real-world datasets, establishing a strong baseline for future RobustPVOS research. Our benchmark is publicly available at https://sohyun-l.github.io/RobustPVOS_project_page/.

Sohyun Lee, Yeho Gwon, Lukas Hoyer +3

POSTECH · Google · ETH Zürich

Video Object Segmentation Objects Peak Memory

May 11, 2026stat.ML

Factual recall in linear associative memories: sharp asymptotics and mechanistic insights

Large language models demonstrate remarkable ability in factual recall, yet the fundamental limits of storing and retrieving input--output associations with neural networks remain unclear. We study these limits in a minimal setting: a linear associative memory that maps

p

input embeddings in

\mathbb{R}^d

to their corresponding~

d

-dimensional targets via a single layer, requiring each mapped input to be well separated from all other targets. Unlike in supervised classification, this strict separation induces~

p

constraints per association and produces strong correlations between constraints that make a direct characterisation of the storage capacity difficult. Here, we provide a precise characterisation of this capacity in the following way. We first introduce a decoupled model in which each input has its own independent set of competing outputs, and provide numerical and analytical evidence that this decoupled model is equivalent to the original model in terms of storage capacity, spectra of the learnt weights, and storage mechanism. Using tools from statistical physics, we show that the decoupled model can store up to

p_c \log p_c / d^2 = 1 / 2

associations, and generalise the computation of

p_c

to linear two-layer architectures. Our analysis also gives mechanistic insight into how the optimal solution improves over a naïve Hebbian learning rule: rather than boosting input-output alignments with broad fluctuations, the optimal solution raises the correct scores just above the extreme-value threshold set by the competing outputs. These findings give a sharp statistical-physics characterisation of factual storage in linear networks and provide a baseline for understanding the memory capacity of more realistic neural architectures.

Alessio Giorlandino, Sebastian Goldt, Antoine Maillard

1International School of Advanced Studies (SISSA), Trieste, Italy · 2INRIA Paris & DI ENS, PSL University, Paris, France

Dense Associative Memory Factual Recall Peak Memory Finite Time Analysis Decoupling

May 11, 2026cs.LG

BROS: Bias-Corrected Randomized Subspaces for Memory-Efficient Single-Loop Bilevel Optimization

Stochastic bilevel optimization (SBO) has become a standard framework for hyperparameter learning, data reweighting, representation learning, and data-mixture optimization in deep learning. Existing exact single-loop SBO methods and memory-efficient surrogate SBO methods either create severe memory pressure for large lower-level neural networks or lack competitive convergence guarantees under standard assumptions. In this paper, we propose BROS, a memory-efficient single-loop SBO method with the same convergence rate order as exact single-loop SBO methods. BROS performs lower and auxiliary updates in randomized subspaces with a Rademacher bi-probe correction that recovers an unbiased Hessian-action estimator. We prove that BROS preserves the

\mathcal O(\varepsilon^{-2})

sample complexity of MA-SOBA for finding an

\varepsilon

-stationary point under only standard assumptions. Experiments on hyper-data cleaning, data-mixture learning, hyper-representation learning, and ViT sample reweighting show that BROS reduces peak memory by up to 44.9% while closely matching full-space baseline performance.

Hengrui Zhang, Boao Kong, Engao Zhang +1

1Sichuan University · 2Peking University

Bilevel Optimization Subspace Peak Memory

May 11, 2026cs.NI

Bridging the Cognitive Gap: A Unified Memory Paradigm for 6G Agentic AI-RAN

As 6G evolves, the radio access network must transcend traditional automation to embrace agentic AI capable of perception, reasoning, and evolution. A fundamental cognitive gap persists in current disaggregated architectures, where interfaces force the physical layer to compress high-dimensional states into low-dimensional metrics, trapping reasoning agents behind a semantic bottleneck. This article envisions a shift from interface-bound to memory-centric architectures. We propose a unified memory paradigm that dissolves the boundaries between sensing and reasoning by mapping biological memory hierarchies onto heterogeneous computing fabrics. Enabled by emerging coherent interconnects, this approach creates a cognitive continuum where microsecond-level reflexes, millisecond-level reasoning, and long-term evolution share state across time scales. By replacing message passing with zero-copy observability, we empower AI agents to bridge the gap between real-time responsiveness and long-horizon context for truly autonomous 6G networks.

Xijun Wang, Zhaoyang Liu, Chenyuan Feng +3

Sun Yat-sen University, China · University of Exeter, U.K. · Zhejiang University, China +2

Open Radio Access Networks Peak Memory

May 10, 2026cs.CR

MemPrivacy: Privacy-Preserving Personalized Memory Management for Edge-Cloud Agents

As LLM-powered agents are increasingly deployed in edge-cloud environments, personalized memory has become a key enabler of long-term adaptation and user-centric interaction. However, cloud-assisted memory management exposes sensitive user information, while existing privacy protection methods typically rely on aggressive masking that removes task-relevant semantics and consequently degrades memory utility and personalization quality. To address this challenge, We propose MemPrivacy, which identifies privacy-sensitive spans on edge devices, replaces them with semantically structured type-aware placeholders for cloud-side memory processing, and restores the original values locally when needed. By decoupling privacy protection from semantic destruction, MemPrivacy minimizes sensitive data exposure while retaining the information required for effective memory formation and retrieval. We also construct MemPrivacy-Bench for systematic evaluation, a dataset covering 200 users and over 155k privacy instances, and introduce a four-level privacy taxonomy for configurable protection policies. Experiments show that MemPrivacy achieves strong performance in privacy information extraction, substantially surpassing strong general-purpose models such as GPT-5.2 and Gemini-3.1-Pro, while also reducing inference latency. Across multiple widely used memory systems, MemPrivacy limits utility loss to within 1.6%, outperforming baseline masking strategies. Overall, MemPrivacy offers an effective balance between privacy protection and personalized memory utility for edge-cloud agents, enabling secure, practical, and user-transparent deployment.

Yining Chen, Jihao Zhao, Bo Tang +5

1MemTensor (Shanghai) Technology Co., Ltd. · 3Tongji University · 2HONOR Device Co., Ltd.

Privacy Peak Memory Edge Platforms Edge Devices

May 9, 2026cs.LG

Sparse Layers are Critical to Scaling Looped Language Models

Looped language models repeat a set of transformer layers through depth, reducing memory costs and providing natural early-exit points at loop boundaries. However, looped models do not scale as favorably as standard transformers with unique layers. We compare standard and Mixture-of-Experts (MoE) transformers, with and without looping, and find two main results. First, we find Looped-MoE models scale better than the standard baseline while dense looped models do not. We trace this to routing divergence between loops: in Looped-MoE models, different experts are activated on each pass through the same shared layers, recovering expressivity without additional parameters. Our second finding is that looped models have better compute-quality trade-offs with early exits than standard models. Because each loop ends with the same layers that produce the final output, loop boundaries are superior exit points, as confirmed by earlier output convergence at these points. In sum, we provide a clear direction for scaling looped models: a Looped-MoE model with early exits can not only beat standard transformers at scale, but also enable significant memory and inference savings with minimal degradation in quality.

Ryan Lee, Jacob Biloki, Edward J. Hu +1

1USC Information Sciences Institute · 2Netflix · 3Independent Researcher

Transformer Architectures Layer-Wise Localized Lora-Moe Mixture-Of-Experts Peak Memory

May 9, 2026cs.LG

VORT: Adaptive Power-Law Memory for NLP Transformers

Standard Transformers impose near-exponential decay on the influence of distant tokens, conflicting with the power-law structure of long-range dependencies in natural language. We introduce the \emph{Variable-Order Retention Transformer} (\VORT{}), a memory architecture in which each ingested token is assigned a learnable fractional order α_i\in[δ,1] that governs a Grünwald--Letnikov power-law retention kernel. Because the fractional weighted sum is non-Markovian, we approximate it through a sum-of-exponentials (SOE) decomposition computed by Gauss--Laguerre quadrature on a Laplace-type integral representation of the kernel weights. Each exponential component admits a one-step Markovian recurrence at O(Sd_v) per step, where S=O(\log(T/\varepsilon)) terms suffice for \varepsilon-uniform accuracy on horizon [1,T]. Retrieval is keyed and associative via a linear-attention accumulator with an exact O(KSd_φd_v) -per-step recurrence. Four results are established: (i) an SOE approximation theorem with geometric convergence rate from the analyticity of the integrand after a log-change of variables; (ii) a quantisation bound valid on [δ,1] with correct analysis near α=0; (iii) a direct L^2 energy argument (Proposition) showing that for α>1/2 any mixture with fixed minimum decay rate Λ>0 incurs L^2([1,T]) error at least N_α(T)-C(Λ)\to\infty, with the Λ-dependence made explicit; and (iv) linear convergence of a gradient plasticity rule under the Polyak--Łojasiewicz condition. Two synthetic experiments confirm the architectural advantage: a Zipf-distributed retrieval benchmark and an entity label-copy task with uniform lag distribution, the latter ruling out prior-matching as an explanation for the power-law kernel's advantage.

Nabil Mlaiki

Department of Mathematics and Sciences, Prince Sultan University, Riyadh, Saudi Arabia

Transformer Architectures Peak Memory

May 8, 2026cs.AI

Human-Inspired Memory Architecture for LLM Agents

Current LLM agents lack principled mechanisms for managing persistent memory across long interaction horizons. We present a biologically-grounded memory architecture comprising six cognitive mechanisms: (1) sleep-phase consolidation, (2) interference-based forgetting, (3) engram maturation, (4) reconsolidation upon retrieval, (5) entity knowledge graphs, and (6) hybrid multi-cue retrieval. Each mechanism addresses a specific failure mode of naive memory accumulation. We introduce a synthetic calibration methodology that derives all pipeline thresholds without benchmark data exposure, eliminating a common source of evaluation leakage. We evaluate on two benchmarks. First, a VSCode issue-tracking dataset (13K issues, 120K events) where deduplication-based consolidation achieves 97.2% retention precision with 58% store reduction (+21.8 pp over baseline). Second, the LongMemEval personal-chat benchmark where we conduct the first streaming M-tier evaluation (475 sessions, ~540K unique turns). At a 200K-token context budget, our pipeline matches raw retrieval accuracy (70.1% vs. 71.2%, overlapping 95% CI) while exposing a tunable accuracy/store-size operating curve. At S-tier scale (50 sessions), dedup-based consolidation yields a +13.3 pp improvement in preference recall.

Doga Kerestecioglu, Alexei Robsky, Clemens Vasters +2

Microsoft

Peak Memory Large Language Model Agents Dialogue Benchmarks Sleep

May 8, 2026cs.LG

Queryable LoRA: Instruction-Regularized Routing Over Shared Low-Rank Update Atoms

We present a data-adaptive method for parameter-efficient fine-tuning of large neural networks. Standard low-rank adaptation methods improve efficiency by restricting each layer update to a fixed low-rank form, but this static parameterization can be too rigid when the appropriate correction depends on the input and on the evolving depth-wise computation of the network. Our approach replaces a purely layer-local adapter with a shared queryable memory of low-rank update atoms. For each block of layers, the model forms a query from the current low-rank state and a running summary of previous blocks, uses this query to retrieve a content-dependent combination of shared update components via attention, and applies the resulting routed operator within the low-rank bottleneck. In this way, the method retains the efficiency and scalability of low-rank adaptation while allowing the effective update to vary across inputs and to share reusable structure across layers. The resulting architecture provides a principled middle ground between static LoRA-style updates and fully generated parameter updates: it remains compact and parameter-efficient while supporting dynamic, context-sensitive adaptation. Further, we incorporate instruction-regularization by augmenting routing logits with a language-induced prior over update atoms, thereby biasing the selection of low-rank transformations toward semantically relevant directions without generating unconstrained parameter updates. Experiments on noisy non-linear regression tasks and LLM fine-tuning suggest that this queryable update-memory formulation can improve final test performance and training stability compared to standard low-rank adaptation, while using a comparable number of trainable parameters.

Omatharv Bharat Vaidya, Connor T. Jerzak, Nhat Ho +1

Department of Computer Science University of Texas at Austin · Department of Government University of Texas at Austin · University of Texas at Austin

Low-Rank Structure Neural Network Peak Memory

May 8, 2026cs.LG

Zero-Shot Imagined Speech Decoding via Imagined-to-Listened MEG Mapping

Decoding imagined speech from non-invasive brain recordings is challenging because imagined datasets are scarce and difficult to align temporally across subjects and sessions In this work, we propose a new approach to the decoding of imagined speech that leverages the richer and more reliably labeled recordings during listening to speech. We collected paired listened and imagined MEG recordings to rhythmic melodic and spoken stimuli from trained musicians. Using trained musicians helped improve temporal alignment across conditions. We then developed a three-stage decoding pipeline that revealed consistent and meaningful relationships between neural activity evoked by imagining and listening to the same stimuli. First, we trained six linear and neural models to map imagined MEG responses to listened responses. We evaluated these models against a null baseline from unseen subjects to validate that the predicted-listening responses preserve stimulus-specific information. In the second stage, we trained a contrastive word decoder exclusively on the listened MEG responses, and evaluated it using four embedding strategies including semantic, acoustic, and phonetic representations. In the third stage, we process the imagined MEG responses from held-out subjects through the mapping pipeline to compute the corresponding listening responses that are then decoded by the listened decoder. Using rank-based analysis, we show that the imagined words are decodable significantly above chance. We shall report here the results of a proof-of-concept implementation to decode imagined speech, where all evaluations are performed on held-out subjects. We also demonstrate that performance improves with training data size, suggesting that this approach is scalable and can directly be made applicable to realistic brain-computer interface scenarios.

Maryam Maghsoudi, Shihab Shamma

University of Maryland · College Park, MD 20740

Neural Recordings Decoding Grapheme-To-Phoneme Peak Memory

May 8, 2026cs.LG

RDKV: Rate-Distortion Bit Allocation for Joint Eviction and Quantization of the KV Cache

Large language models (LLMs) have shown strong performance across diverse tasks, but their inference with long input contexts is bottlenecked by memory size and bandwidth. The Key-Value (KV) cache size grows linearly with sequence length and needs to be re-read from off-chip high-bandwidth memory (HBM) to on-chip memory at every decoding step, resulting in memory-bound inference. Existing methods reduce the cache by either eviction or quantization, but typically treat the two in isolation. In this paper, we cast KV cache compression as a rate-distortion problem, under which eviction and quantization are two end-points of the same bit allocation scheme. This exposes the need to optimize them jointly, motivating our method, RDKV (Rate-Distortion KV cache compression). RDKV derives the weight of each token or channel from the distortion that compression induces on the attention computation. Based on these weights, it assigns each token or channel a bit-width ranging from full precision down to zero bits guided by reverse water-filling, applied once after the prefilling stage. Experiments on LongBench, RULER, and InfiniteBench show that RDKV outperforms the best evaluated baseline by 9.1% on average. On LongBench it recovers 97.81% of full-cache accuracy with only 2.48% cache retention. Compared with full-cache FlashAttention-2 decoding, it achieves 4.5x decode speedup and 1.9x peak memory reduction with 128K context length, while maintaining comparable performance.

Junkai Zhang, Hang Guo, Luca Benini +1

ETH Zurich · Tsinghua University

Key-Value Cache Compression Key-Value Cache Rate-Distortion Theory Peak Memory Context Length

May 8, 2026cs.LG

mHC-SSM: Manifold-Constrained Hyper-Connections for State Space Language Models with Stream-Specialized Adapters

Manifold-Constrained Hyper-Connections (mHC) introduce a stability-motivated variant of multi stream residual mixing by constraining residual stream mixing matrices to the manifold of doubly stochastic matrices via Sinkhorn-Knopp projection. In his work, we study whether mHC-style constrained multi-stream residual topology transfers effectively to state space model (SSM) language modeling. We implement a static mHC mechanism around an SSM block by expanding the residual stream into multiple parallel streams, aggregating streams into a single SSM input through simplex-constrained pre-mixing, scattering the SSM output back to streams through simplex-constrained post-mixing, and applying Sinkhorn-projected residual stream mixing at each layer. We further introduce stream-specialized adapters that add lightweight stream-specific capacity through a shared bottleneck with per-stream scaling, applied both before stream aggregation and after the SSM output prior to scattering. We evaluate baseline single-stream SSM, static mHC SSM, and mHC SSM with adapters on WikiText-2 using identical training settings and report checkpoint-based validation loss, perplexity, throughput, and peak GPU memory. Under the reported fair checkpoint evaluation, static mHC improves validation loss from 6.3507 to 6.2448 and reduces perplexity from 572.91 to 515.35, while mHC with adapters further improves validation loss to 6.1353 and perplexity to 461.88. These gains are accompanied by modest throughput reductions from 1025.52 to 964.81 and 938.90 tokens per second, and increased peak memory from 2365 MB to 2568 MB and 3092 MB. The results suggest that mHC-inspired constrained multi-stream residual mixing can yield measurable quality improvements in SSM language models and that stream-specialized adapter capacity can further enhance performance with predictable efficiency tradeoffs.

Abdulvahap Mutlu, Şengül Doğan, Türker Tuncer

Department of Digital Forensics Engineering, Technology Faculty, Firat University, Elazig, Turkey

May 8, 2026cs.CV

Bridging Modalities, Spanning Time: Structured Memory for Ultra-Long Agentic Video Reasoning

Understanding ultra-long videos such as egocentric recordings, live streams, or surveillance footage spanning days to weeks, remains a challenge. For current multimodal LLMs: even with million-token context windows, frame budgets cover only tens of minutes of densely sampled video, and most evidence is discarded before inference begins. Memory-augmented and agentic approaches help with scale, but their retrieval remains fragmented across modalities and lacks long-range narrative summaries that span days or weeks. We propose \textbf{MAGIC-Video}, a training-free framework built around a multimodal memory graph with interleaved narrative chain: the graph unifies episodic, semantic, and visual content through six typed edges and supports cross-modal retrieval, while the chain distils long-horizon entity biographies and recurring activity events. At inference time, an agentic loop interleaves graph retrieval with narrative fact injection, covering both the modality and time dimensions of ultra-long video in a single retrieval pipeline. On EgoLifeQA, Ego-R1 and MM-Lifelong, MAGIC-Video consistently outperforms strong general-purpose, long-video, and agentic baselines, with gains of 10.1, 7.4, and 5.9 points over the prior best agentic system on each benchmark. Code is available at https://github.com/lijiazheng0917/MAGIC-video.

Jiazheng Li, Chi-Hao Wu, Yunze Liu +3

University of Connecticut · Northwestern University · University of Virginia

Video Agent Multimodal Memory Graph-Structured Memories Agent Loop Complementary Modalities Peak Memory

May 7, 2026cs.LG

How to Compress KV Cache in RL Post-Training? Shadow Mask Distillation for Memory-Efficient Alignment

Reinforcement Learning (RL) has emerged as a crucial paradigm for unlocking the advanced reasoning capabilities of Large Language Models (LLMs), encompassing frameworks like RLHF and RLAIF. Regardless of the specific optimization algorithm (e.g., PPO, GRPO, or Online DPO), online RL inherently requires an exploratory trajectory generation (rollout) phase. However, for long-context reasoning tasks, this rollout phase imposes a severe ``memory wall'' due to the exorbitant Key-Value (KV) cache footprint. While applying KV cache compression during rollouts mitigates this memory overhead, it induces a critical off-policy bias. Although modern KV compression is often nearly lossless during standard inference, even minuscule approximation errors are drastically amplified by the inherent instability of RL optimization. Specifically, the sampler generates responses under a sparse context, whereas the learner updates parameters using the full, dense context. Existing statistical solutions, such as importance reweighting, struggle to correct this magnified bias, suffering from high gradient variance and severe sample inefficiency.

Rui Zhu, Weiheng Bai, Qiushi Wu +3

Yale University · University of Minnesota Twin Cities · Indiana University Bloomington

Key-Value Cache Compression Reinforcement Learning Post-Training Key-Value Cache Peak Memory Few-Step Distillation

May 7, 2026cs.AI

Belief Memory: Agent Memory Under Partial Observability

LLM agents that operate over long context depend on external memory to accumulate knowledge over time. However, existing methods typically store each observation as a single deterministic conclusion (e.g., inferring "API~X failed" from temporary errors), even though such observations are inherently partial and potentially ambiguous. By committing to one conclusion and discarding uncertainty, these methods introduce self-reinforcing error: the agent acts on the stored conclusion, never revisits alternatives, and reinforces the conclusion over time. To address this issue, we propose BeliefMem, which shifts the memory paradigm from committing to a single conclusion per observation to retaining multiple candidate conclusions with their probabilities. Concretely, BeliefMem stores the candidate conclusions as separate memory entries, each carrying a probability that is updated via Noisy-OR rules as new observations arrive. At retrieval, all candidates surface together with their probabilities, keeping alternatives visible to the agent. Since each conclusion in memory retains its probability, BeliefMem preserves the uncertainty that the deterministic paradigm discards, enabling the agent to act with high confidence on well-evidenced knowledge while retaining the capacity to update its confidence when new evidence arrives. Empirical evaluations on LoCoMo and ALFWorld benchmarks show that, even with limited data, BeliefMem achieves the best average performance, remarkably outperforming well-known baselines. More broadly, such probabilistic memory produces substantial gains and explores a new direction for agent memory in partially observable environments.

Junfeng Liao, Qizhou Wang, Jianing Zhu +3

MBZUAI · RIKEN AIP · UT Austin +1

Long-Term Agent Memory Partially Observable Markov Decision Process Peak Memory

May 6, 2026cs.CL

Storage Is Not Memory: A Retrieval-Centered Architecture for Agent Recall

Extraction at ingestion is the wrong primitive for agent memory: content discarded before the query is known cannot be recovered at retrieval time. We propose True Memory, a six-layer architecture that shifts the center of the system from a storage schema to a multi-stage retrieval pipeline operating over events preserved verbatim. The full system runs as a single SQLite file on commodity CPU with no external database, vector index, graph store, or GPU. On LoCoMo (1,540 questions across 10 multi-session conversations), True Memory Pro reaches 93.0% accuracy (3-run mean) against 61.4% for Mem0, 65.4% for Supermemory, approximately 71% for Zep, and 94.5% for EverMemOS under a matched gpt-4.1-mini answer model. On LongMemEval (500 questions), True Memory Pro reaches 87.8% (3-run mean). On BEAM-1M (700 questions at the 1-million-token scale), True Memory Pro reaches 76.6% (3-run mean), above the prior published result of 73.9% for Hindsight. A 56-configuration ablation shows a 1.3-percentage-point spread within the top-performing configuration family.

Joshua Adler, Guy Zehavi

Sauron Labs

Factual Recall Peak Memory Retrievers