Collapse

Follow this field Ask about this field

Recent momentum

emerging

0 papers in the last 28 days · 0.0% of indexed attention

Twelve weeks of publication activity for this topic as it is defined today.

Weekly history

Recent digests

What was published in this field, kept on the site without email delivery.

Period ending 2026-09-21

3 new papers

A weekly snapshot of new work published in Collapse.

Period ending 2026-09-14

6 new papers

A weekly snapshot of new work published in Collapse.

Period ending 2026-09-07

5 new papers

A weekly snapshot of new work published in Collapse.

Inside this field

Focused directions

Focused topic

Collapse

Within Collapse

100 papersExplore

Focused topic

Representation Collapse

Within Collapse

41 papersExplore

Focused topic

Entropy Collapse

Within Collapse

28 papersExplore

Focused topic

Neural Collapse

Within Collapse

27 papersExplore

181 papers

Latest in Collapse

Browse all topics →

Multi-LLM Systems Exhibit Robust Semantic Collapse
May 16, 2026Weiyi Kong, Shiyang Lai, Jinghua Piao +1Collapse Counter Turing Test
The Geometry of Projection Heads: Conditioning, Invariance, and Collapse
May 16, 2026Faris ChaudhryMulti-Head Self-Attention Riemannian Manifolds
Provably Learning Diffusion Models under the Manifold Hypothesis: Collapse and Refine
May 16, 2026Wei Huang, Andi Han, Mingyuan Bai +4Score-Based Diffusion Model Discrete Diffusion Model
Conflict-Free Replicated Data Types for Neural Network Model Merging: A Two-Layer Architecture Enabling CRDT-Compliant Model Merging Across 26 Strategies
May 16, 2026Ryan GillespieModel Merging Replication
Entropic Auto-Encoding via Implicit Free-Energy Minimization
May 15, 2026Hazhir Aliahmadi, Irina Babayan, Greg van AndersVariational Autoencoder Masked Autoencoders
XDomainBench: Diagnosing Reasoning Collapse in High-Dimensional Scientific Knowledge Composition
May 14, 2026Gong Zhiren, Tiantong Wu, Jiaming Zhang +9Scientific Discovery Failure Mechanism
Silent Collapse in Recursive Learning Systems
May 14, 2026Zhipeng ZhangCollapse Trajectory-Level Credit
MahaVar: OOD Detection via Class-wise Mahalanobis Distance Variance under Neural Collapse
May 14, 2026Donghwan Kim, Hyunsoo YoonOut-Of-Distribution Detection In-Distribution
Reliability-Gated Source Anchoring for Continual Test-Time Adaptation
May 13, 2026Vikash Singh, Debargha Ganguly, Weicong Chen +8Continual Test-Time Adaptation Independent Anchor Network
Controlling Logical Collapse in LLMs via Algebraic Ontology Projection over F2
May 13, 2026Hisashi Miyashita, Mgnite IncOntology Collapse
Persona-Model Collapse in Emergent Misalignment
May 13, 2026Davi Bastos Costa, Renato VicenteMisalignment Moral Reasoning
Cluster-Aware Neural Collapse Prompt Tuning for Long-Tailed Generalization of Vision-Language Models
May 12, 2026Boyang Guo, Liang Li, Lin Peng +3Prompt Tuning Large Vision Language Models
Beyond Point-wise Neural Collapse: A Topology-Aware Hierarchical Classifier for Class-Incremental Learning
May 12, 2026Huiyu Yi, Zhiming Xu, Dunwei Tu +3Domain Incremental Learning Neural Collapse
More Edits, More Stable: Understanding the Lifelong Normalization in Sequential Model Editing
May 12, 2026Xin Ma, Wei Chen, Qi Liu +4Multimodal Knowledge Editing Catastrophic Forgetting
Understanding and Preventing Entropy Collapse in RLVR with On-Policy Entropy Flow Optimization
May 12, 2026Huimin Xu, Shuai Zhao, Xiaobao Wu +1Reinforcement Learning With Verifiable Reward Flow-Based Policy
Optimal Representations for Generalized Contrastive Learning with Imbalanced Datasets
May 11, 2026Thuan Nguyen, Shuchin Aeron, D. Richard Brown +1Contrastive Learning Class Imbalance
Rethinking Loss Reweighting for Imbalance Learning as an Inverse Problem: A Neural Collapse Point of View
May 11, 2026Jinping Wang, Zixin Tong, Zhiwu Xie +1Loss Function Neural Collapse
Flag Varieties: A Geometric Framework for Deep Network Alignment
May 11, 2026Jingchuan Xiao, Xinyi Sui, Cihan RuanNeural Collapse Orthogonality
Path-Dependent Denoising: A Non-Conservative Field Perspective on Order Collapse in Diffusion Language Models
May 10, 2026Jeonseong KimDiffusion Language Models Image Denoising
How You Begin is How You Reason: Driving Exploration in RLVR via Prefix-Tuned Priors
May 9, 2026Yifan Xu, Junren Chen, Yifan ChenReinforcement Learning With Verifiable Reward Verifiable Rewards
Anchoring the Eigengap: Cross-Modal Spectral Stabilization for Sample-Efficient Representation Learning
May 9, 2026Nikhil J. Dhinagar, Vidhi Chhatbar, Chirag Jagad +6Representation Learning Cross-Modal Learning
Curated Synthetic Data Doesn't Have to Collapse: A Theoretical Study of Generative Retraining with Pluralistic Preferences
May 8, 2026Ali Falahati, Mohammad Mohammadi Amiri, Kate Larson +1Synthetic Data Generative Models
Continuous First, Discrete Later: VQ-VAEs Without Dimensional Collapse
May 7, 2026Xinyu Zhao, Nikita Karagodin, Hamed Hassani +3Variational Autoencoder Codebooks
Layer Collapse in Diffusion Language Models
May 7, 2026Alexander Conzelmann, Albert Catalan-Tatjer, Shiwei LiuDiffusion Language Models Large Language Model Compression
The Weight Gram Matrix Captures Sequential Feature Linearization in Deep Networks
May 7, 2026Taehun Cha, Daniel Beaglehole, Adityanarayanan Radhakrishnan +1Feature Learning Deep Network
Mean Mode Screaming: Mean--Variance Split Residuals for 1000-Layer Diffusion Transformers
May 7, 2026Pengqi LuMultimodal Diffusion Transformers Transformer Residual Streams
A Testable Certificate for Constant Collapse in Teacher-Guided VAEs
May 7, 2026Zegu Zhang, Jianhua Peng, Jian ZhangEntropy Collapse Collapse
On Semantic Loss Fine-Tuning Approach for Preventing Model Collapse in Causal Reasoning
May 6, 2026Pratik Deshmukh, Atirek GuptaCausal Reasoning Collapse
Paraphrase-Induced Output-Mode Collapse: When LLMs Break Character Under Semantically Equivalent Inputs
May 6, 2026Aofan Liu, Jingxiang MengLarge Language Model Performance Representation Collapse
Position: the Stochastic Parrot in the Coal Mine. Model Collapse is a Threat to Low-Resource Communities
May 5, 2026Devon Jarvis, Richard Klein, Benjamin Rosman +2Collapse Bird
Private Speech Classification without Collapse: Stabilized DP Training and Offline Distillation
May 4, 2026Yadi Wen, Tianxin Li, Enji Liang +2Speaker Recognition Efficient Privacy Loss
When Attention Collapses: Residual Evidence Modeling for Compositional Inference
May 4, 2026Niklas HoubaIncomplete Evidence Representation Collapse
ANO: Robust Policy Optimization via Bounded, Redescending Gain Fields
May 4, 2026Yiheng Zhang, Yiming Wang, Kaiyan Zhao +3Proximal Policy Optimization Multi-Objective Policy Optimization
How Label Imbalance Shapes Geometry: A General Spectral Analysis of Multi-Label Neural Collapse
May 3, 2026Xiaoxuan Ma, Yixuan Yang, Song Li +1Multi-Label Classification Neural Collapse
Recall to Predict: Grounding Motion Forecasting in Interpretable Motion Bank
May 2, 2026Abhishek Vivekanandan, Ahmed Abouelazm, J. Marius ZöllnerHuman Motion Prediction Factual Recall
EmoMM: Benchmarking and Steering MLLM for Multimodal Emotion Recognition under Conflict and Missingness
May 1, 2026Yueru Sun, Yimeng Zhang, Haoyu Gu +5Emotion Recognition Modality Gap
Posterior Augmented Flow Matching
May 1, 2026George Stoica, Sayak Paul, Matthew Wallingford +6Hierarchical Flow Matching Collapse
Escaping Mode Collapse in LLM Generation via Geometric Regulation
May 1, 2026Xin Du, Kumiko Tanaka-IshiiLarge Language Model Generation Autoregressive Generation
Why Self-Supervised Encoders Want to Be Normal
Apr 30, 2026Yuval DombInformation Bottleneck Latent Representation
Why Mean Pooling Works: Quantifying Second-Order Collapse in Text Embeddings
Apr 30, 2026Tomomasa Hara, Hiroto Kurita, Masaaki Imaizumi +2Domain-Invariant Text Embeddings Global Embeddings
Instruction Complexity Induces Positional Collapse in Adversarial LLM Evaluation
Apr 29, 2026Jon-Paul CacioliLanguage Model Evasion Attacks Collapse
Exploring the Limits of Pruning: Task-Specific Neurons, Model Collapse, and Recovery in Task-Specific Large Language Models
Apr 29, 2026M. K. Khalidi Siam, Md. Tausif-Ul-Islam, Md. Reshad Romim Khan +5Pruning Selective
Addressing Performance Saturation for LLM RL via Precise Entropy Curve Control
Apr 29, 2026Bolian Li, Yifan Wang, Yi Ding +3Reinforcement Learning Post-Training Entropy Collapse
Monitoring Neural Training with Topology: A Footprint-Predictable Collapse Index
Apr 28, 2026Alexander KalinowskiNeural Collapse Topological Data Analysis
The Chameleon's Limit: Investigating Persona Collapse and Homogenization in Large Language Models
Apr 27, 2026Yunze Xiao, Vivienne J. Zhang, Chenghao Yang +3Personality Large Language Model Performance
Quantum Kernel Advantage over Classical Collapse in Medical Foundation Model Embeddings
Apr 27, 2026Sebastian Cajas Ordóñez, Felipe Ocampo Osorio, Dax Enshan Koh +10Quantum Kernels Medical Image Classification
The Collapse of Heterogeneity in Silicon Philosophers
Apr 26, 2026Yuanming Shi, Andreas HauptPhilosophy Artificial Intelligence Alignment
Pairing Regularization for Mitigating Many-to-One Collapse in GANs
Apr 22, 2026Kuan-Yu Lin, Yu-Chih Huang, Tie LiuGenerative Adversarial Network Regularization Techniques
A Delta-Aware Orchestration Framework for Scalable Multi-Agent Edge Computing
Apr 22, 2026Samaresh Kumar Singh, Joyjit RoyHeterogeneous Edge Devices Multi-Agent Orchestration Frameworks
Domain-Specialized Object Detection via Model-Level Mixtures of Experts
Apr 20, 2026Svetlana Pavlitska, Malte Stüven, Beyza Keskin +1Object Detection Mixtures
Neural Garbage Collection: Learning to Forget while Learning to Reason
Apr 20, 2026Michael Y. Li, Jubayer Ibn Hamid, Emily B. Fox +1Catastrophic Forgetting Large Language Model Memory
HEALing Entropy Collapse: Enhancing Exploration in Few-Shot RLVR via Hybrid-Domain Entropy Dynamics Alignment
Apr 20, 2026Zhanyu Liu, Qingguo Hu, Ante Wang +5Reinforcement Learning With Verifiable Reward Verifiable Rewards
Language Generation with Replay: A Learning-Theoretic View of Model Collapse
Mar 12, 2026Giorgio Racca, Michal Valko, Amartya SanyalCollapse
Do Implicit Personalization and Explicit Styles Conflict? PsPLUG: A Lightweight Plug-in for Balancing Personalization and Style in Customized LLMs
Jan 10, 2026Yutong Song, Jiang Wu, Shaofan Yuan +5Personalization Model Preference
Escaping Model Collapse via Synthetic Data Verification: Near-term Improvements and Long-term Convergence
Oct 18, 2025Bingji Yi, Qiyuan Liu, Yuwei Cheng +1Synthetic Data Retraining
Verbalized Sampling: How to Mitigate Mode Collapse and Unlock LLM Diversity
Oct 1, 2025Jiayi Zhang, Simon Yu, Derek Chong +4Diversity Large Language
GeLaCo: An Evolutionary Approach to Layer Compression
Jul 14, 2025David Ponce, Thierry Etchegoyhen, Javier Del SerLarge Language Model Compression Context Compression
Recursive Learning Without Collapse: A Weighting-Based Stabilization Framework
Feb 25, 2025Hengzhi He, Shirong Xu, Guang ChengSynthetic Data Generative Models
You Do Not Fully Utilize Transformer's Representation Capacity
Feb 13, 2025Gleb Gerasimov, Yaroslav Aksenov, Nikita Balagansky +2Transformer Layers Transformer Architectures
Learning by Surprise: Adaptive Mitigation of Model Collapse in Large Language Models
Oct 16, 2024Daniele Gambetta, Gizem Gezici, Fosca Giannotti +3Collapse Generative Artificial Intelligence

May 16, 2026 · cs.MAJ/K move · Enter open · S save

Multi-LLM Systems Exhibit Robust Semantic Collapse

Weiyi Kong, Shiyang Lai, Jinghua Piao, James Evans

Department of Electrical Computer Engineering, University of Toronto; Toronto, M5S 3G4, Canada · Department of Sociology, University of Chicago; Chicago, 60637, USA · 3Knowledge Lab, University of Chicago; Chicago, 60637, USA · Department of Electronic Engineering, Tsinghua University; Beijing, 100084, China · 5Google, Paradigms of Intelligence Team; Mountain View, 94043, USA

Whether machines can originate novel content has been debated for nearly two centuries, from Lovelace's assertion that no engine can "originate anything" to Turing's question of whether a machine can amplify ideas brought in from outside. Systems of multiple interacting LLMs, increasingly deployed for autonomous generation, reopen this question empirically. Here we show that such systems, operating in inference-only setups, exhibit semantic collapse: systematic convergence in semantic representations despite apparent lexical variation. Across model families, extended simulations of 200 to 1,000 rounds, the pattern remains consistent. Thirteen intervention strategies, spanning decoding parameters, prompt design, agent composition, activation engineering, and reinforcement learning, fail to restore semantic diversity. Mechanistic analyses suggest that semantic collapse is not explained by alignment or conformity biases, but is consistent with intrinsic properties of autoregressive generation. Our results point to persistent constraints on the ability of multi-LLM systems to sustain open-ended exploration in closed-loop settings.

Collapse Counter Turing Test Convergence

Open paper Ask about it PDF