Adaptive Computation

Follow this topic Ask about this topic

Adaptive computation studies how models can spend more processing on difficult or uncertain inputs and skip redundant work on easier ones, improving efficiency without giving up accuracy. Common approaches include early-exit networks, token routing, recurrent and latent reasoning with learned halting, and difficulty- or uncertainty-based compute routers; related methods adapt visual observation and planning budgets. These techniques are tested in settings such as mathematical reasoning, language modeling, multimodal video, edge vision, and robotics, with efficiency measured in FLOPs, tokens, latency, memory, or energy. A central challenge is making predicted compute needs reliable while ensuring savings translate into real speedups, and evaluating gains against strong fixed-compute baselines across changing tasks and hardware.

Latest papers

103 papers23 papers in the last four weeks, up 44% on the four weeks before. 0.1% of all new papers.

DEQ-Stereo: Deep Euilibrium Video Stereo Matching via Adaptive Convergence
Oct 29, 2026Kiljoon Han, Kyumin Hwang, Wonheok Choi +2Stereo Matching Adaptive Computation
Daegu Gyeongbuk Institute of Science and Technology, Daegu, South Korea · Korea Advanced Institute of Science and Technology, Daejeon, South Korea
Does Scaling Toward AGI Depend on What System-1 Lets System-2 Execute?
Oct 29, 2026Qi Wang, Yun Qu, Lizhou Cai +10AI Agent Architecture LLM Agent Efficiency
Department of Automation, Tsinghua University, Beijing, China · Department of Computer Science, Tsinghua University, Beijing, China
Beyond Causal Masks: Auditing and Localizing Hidden Future-Token Leakage in Adaptive Sequence Models
Oct 22, 2026Huiting Liu, Jinsong WuLook-Ahead Bias in Language Models LLM Auditing
Department of Electrical Engineering, University of Chile, Santiago, 8370451, Chile
Think Harder Where It Matters: Item-Level Adaptive Recursion for Sequential Recommendation
Oct 21, 2026Zhanshuo Cao, Jiakai Tang, Yuqi Zhou +2Sequential Recommendation Adaptive Computation
Gaoling School of Artificial Intelligence, Renmin University of China, Beijing, China
Adaptive Replay Refinement: Allocating Test-Time Compute Between Revision and Resampling
Oct 20, 2026Ahsan Bilal, Muhammad Ahmed Mohsin, Muhammad Umer +3Inference-Time Compute Allocation Test-Time Scaling
University of Oklahoma · Stanford University · National University of Sciences and Technology
recAMPLIFY: Large-Scale Pretraining of Recursive Protein Language Models
Oct 19, 2026Saurav Jha, Qianli Liao, Robert Vernon +2Protein Language Models Recursive Language Models
Mila – Quebec AI Institute · Polytechnique Montréal · Center for Brains, Minds, and Machines +4
Sep-Latent: Separator Tokens as Natural Compression Anchors for Implicit Reasoning
Oct 16, 2026Yinlong Xu, Jianfeng Huang, Zhengdong Zhou +5Efficient Language Model Reasoning Implicit Reasoning in Language Models
College of Computer Science and Technology, Zhejiang University, Hangzhou, China · Alibaba Group, Hangzhou, China
LEV: Thinking Fast and Deep with Looped Decision Models
Oct 14, 2026Dingjie Song, Zhiling Yan, Yuxuan Zhang +1Efficient Language Model Inference Adaptive Computation
Lehigh University · University of British Columbia · Vector Institute
Adaptive Observation Control for Active Visual Recognition
Oct 14, 2026Dikai Li, Weijia Wang, Yuan-Gen WangActive Visual Perception Adaptive Computation
School of Artificial Intelligence, Guangzhou University, Guangzhou, China
Adaptive Computation with Conformal Certification
Oct 13, 2026Shehtab Zaman, Aneesh Rangnekar, Nancy Lee +1Adaptive Computation Conformal Risk Control
Department of Medical Physics, Memorial Sloan Kettering Cancer Center, New York, NY, USA · Department of Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, NY, USA
Resample the Answer: Where to Spend Test-Time Compute in Recursive Reasoning
Oct 12, 2026Maxime IllouzRecursive Reasoning Models Inference-Time Search
Concordia University Mila – Québec AI Institute Montréal, Canada
SpikCTU: A Compact Spiking Network with Adaptive Recurrent Computation for Speech Command Recognition
Oct 10, 2026Li Jin, Jiang Wenbin, Chen Tao +1Spiking Neural Networks Efficient Neural Network Inference
Hangzhou Dianzi University · Olympus Hedge Fund Investment Co., Ltd. · Shanghai Jiao Tong University
Phonology-Aware Speaker Extraction Frontend for Audio LLMs
Oct 9, 2026Yanxiang Chen, Haoyuan Deng, Jinyu Wang +4Target Speaker Extraction Speech Processing
Sun Yat-Sen University, Zhuhai, China · Beijing Normal-Hong Kong Baptist University, Zhuhai, China · Fudan University, Shanghai, China
One Model to Loop Them All: Elastic Latent Recurrence for Language Models
Oct 8, 2026Faisal Ladhak, Shantanu Acharya, Santiago Akle Serano +3Language Modeling Recurrent Language Models
NVIDIA
Easy to anticipate, hard to compute: boundary dependence finds the computed outputs that entropy patching misses
Oct 8, 2026Nicolás Vera ZúñigaByte-Level Language Model Efficient Language Model Inference
Independent Researcher, Chile
Input-Blind Controls Produce Substantial Oracle Headroom for Layer Programs in Multiple-Choice Evaluation
Oct 7, 2026Yibei Guo, Rui LiuLayer Skipping in Language Model Decoding Input Order Sensitivity in Language Models
College of Aeronautics and Engineering, Kent State University
Enabling Dynamic Computation in Looped LMs
Oct 6, 2026Aayush Mishra, Arnau Padrés Masdemont, Victor Conchello Vendrell +3LLM Inference Acceleration Adaptive Computation
Qualcomm AI Research · Johns Hopkins University
AdaLoop: Adaptive-Depth Latent Reasoning for Audio Language Models
Oct 3, 2026Lee Seung-woo, Bowen QiAudio-Language Models Audio Reasoning
Seoul National University · Shanghai Jiao Tong University
BranchIP: Learning Adaptive Equivariant Computation for Interatomic Potentials
Oct 1, 2026Laura Zichi, Gil Harari, Chuin Wei Tan +6Machine Learning Interatomic Potentials Equivariant Neural Networks
Harvard University · Robert Bosch LLC Research and Technology Center
A Tilted Bowl Is Not a Slippery Slope: Compressing Looped Models
Sep 30, 2026Steven Kolawole, Pearse Jim, Opegbemi M. Busoye +2Efficient Neural Network Inference Recurrent Neural Networks
Carnegie Mellon University · ML Collective
DeepJEPA: Scaling World Models from Within
Sep 30, 2026Zijian Jin, Yunbei Zhang, Yuanzhe Liu +5World Model Learning World Model-Based Planning
New York University · Tulane University · UIUC +4
AdaOcc: Adaptive 3D Occupancy Prediction for Embodied Tasks
Sep 30, 2026Jinglong Wang, Yunjie Wang, Zhiyang Zhang +43D Occupancy Prediction Adaptive Computation
Beihang University · Beijing Academy of Artificial Intelligence · Hebei University of Technology +3
LoopICL: Looping a single transformer block to solve tabular tasks
Sep 28, 2026Amir Rezaei Balef, Katharina EggenspergerTabular Foundation Models Cost-Aware Inference
TU Dortmund University, Dortmund, Germany · Lamarr Institute for Machine Learning and Artificial Intelligence, Dortmund, Germany · University of Tübingen, Tübingen, Germany
Improving Test-Time Scaling with Adaptive Looped Transformers
Sep 28, 2026Yichen You, Tianyu Fu, Aosong Feng +4Test-Time Scaling Recurrent Transformers
Tsinghua University · Yale University
GenoMorph: Pathway-Grounded Genomic Disease Reasoning via Adaptive Latent Computation
Sep 28, 2026Tanmoy Kanti Halder, Akash Ghosh, Arijit Roy +1LLM Reasoning with Graphs Genomics
Department of Computer Science and Engineering, Indian Institute of Technology Patna, Bihta, 801106, Bihar, India · Prasannadeb Women’s College, India
MoR-MLLM: Mixture of Recursions for Efficient Multimodal Large Language Models
Sep 27, 2026Pengcheng Zheng, Chaoning Zhang, Jiaxin Yan +8Multimodal Large Language Models Efficient VLM Inference
Center for Future Media and School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu 611731, China · School of Computing, Kyung Hee University, Yongin-si, South Korea · Nota Inc., Seoul, South Korea +1
DPAMixerSR: An Efficient Degradation-Pattern-Aware Model for Image Super-Resolution
Sep 26, 2026Song-Li Wu, Haonan Jiang, Jixuan Fan +3Image Super-Resolution Adaptive Computation
Shenzhen International Graduate School, Tsinghua University
Certainty Is Not Just Correctness: Rethinking Token-Level Certainty in LLM Reasoning
Sep 26, 2026Yunfan Zhou, Ye Zhu, Zhihai Wang +3Confidence Estimation in Language Models LLM Uncertainty Estimation
School of Computer Science, Shanghai Jiao Tong University · Qwen Team, Alibaba Group
Programs-of-Layers in LLMs through the Lens of Cortical Areas
Sep 25, 2026Justus Westerhoff, Stephan Olbrich, Hatem Oraby +2Transformer Inference Adaptive Computation
Berliner Hochschule für Technik (BHT), Berlin, Germany · Humboldt-Universität zu Berlin, Berlin, Germany
Selective Amortization of Full-Budget Counterfactual Reasoning for Visual Token Communication
Sep 25, 2026Qinglei Qi, Zhihe Liang, Fengzhan Jing +5Generative Image Compression Adaptive Computation
School of Artificial Intelligence, Nanyang Normal University, Nanyang, China · School of Computer and Information Engineering, Tianjin Normal University, Tianjin, China · School of Information Science and Engineering, Linyi University, Linyi 276000, China

Oct 29, 2026J/K move · Enter open · S save

DEQ-Stereo: Deep Euilibrium Video Stereo Matching via Adaptive Convergence

Kiljoon Han, Kyumin Hwang, Wonheok Choi, Jeonghoon Kim+1

Daegu Gyeongbuk Institute of Science and Technology, Daegu, South Korea · Korea Advanced Institute of Science and Technology, Daejeon, South Korea

Stereo Matching Adaptive Computation

Open paper AskPDF arXiv

Iterative refinement models have significantly improved stereo matching accuracy, but their rigid, fixed update schedules incur heavy computational costs and a strict trade-off between speed and accuracy. This challenge is magnified in video streams, where traditional online methods resort to reducing feature channels or relying on expensive temporal warping. To address this, we propose DEQ-Stereo, a novel video stereo matching framework based on the Deep Equilibrium (DEQ) paradigm. Instead of unrolling a fixed number of iterations, DEQ-Stereo reformulates the refinement as an implicit fixed-point problem, adaptively terminating once convergence is met. To ensure fast and stable intra-frame convergence, we introduce the Convolutional Simplified Gated Unit (ConvSGU), which eliminates unnecessary non-linearities by modeling the latent state as a gated, element-wise sum of inputs. Furthermore, to exploit cross-frame temporal coherence, we propose Temporal State Transfer (TST), a warm-start strategy that propagates the converged hidden state to the next frame without requiring explicit pose or warping priors. Extensive evaluations demonstrate that DEQ-Stereo delivers state-of-the-art performance with a superior accuracy-to-latency trade-off.