Language Model Robustness

Follow this topic Ask about this topic

Momentum

35 papers in the last four weeks, up 46% on the four weeks before. 0.3% of all new papers.

Jul 13Week of Sep 28

Latest papers 338

All topics

The Weight of Silence: A Causal Case for Weights Over the Scratchpad in Latent Chess Reasoning
Jul 23, 2026Ishan S. KshirsagarRL for Language Model Reasoning Continuous Latent Reasoning
Independent Researcher
Code Monitor Red Teaming for Public-Test-Passing Code
Jul 23, 2026Junchi Liao, Jiawen Deng, Fuji RenLLM Auditing Code Generation Evaluation
University of Electronic Science and Technology of China Chengdu, China
Understanding the Impact of Linguistic Realization Choices on LLM Stance with Causal Tracing
Jul 22, 2026Langchen Huang, Sebastian Padó, Franziska WeeberPrompt Sensitivity LLM Interpretability
Institute for Natural Language Processing, University of Stuttgart
It's Not What You Say, It's How You Say It: Evaluating LLM Responses to Expressions of Belief
Jul 20, 2026Kevin Du, Clara Kümpel, Michelle Wastl +1LLM Evaluation Knowledge Conflicts in Language Models
ETH Zürich · University of Zurich · UC San Diego
Logical Judgments Under Pressure: Diagnosing Syllogistic Stability with Learned Soft Prefixes
Jul 20, 2026Brian K ChenLogical Reasoning Symbolic Reasoning
National University of Singapore
StabilityBench: Benchmarking Instability in LLMs
Jul 17, 2026Emma Kondrup, Zachary Yang, Anne Imouza +1LLM Evaluation Language Model Generation Evaluation
1Mila — Quebec AI Institute · 2McGill University · 3Ubisoft La Forge
Conditional Reliability of Toxicity Signals for Multilingual and Code-Mixed Abuse Detection
Jul 17, 2026Indraveni Chebolu, Rohan Singh, Arnab Mallick +1Multilingual Language Model Evaluation Hate Speech Detection
Centre for Development of Advanced Computing, Hyderabad, India
Show Me How You Reason and I'll Tell You Who You Are: Reasoning Graphs for Robust LLM Authorship Attribution
Jul 16, 2026Zlata Kikteva, Artur Romazanov, Annette Hautli-Janisz +1AI-Generated Text Detection Authorship Attribution
University of Passau, Germany · University of Dundee, United Kingdom
Routing Ceilings Are Domain-Independent: Structural Prior Injection in Code Security Vulnerability Detection
Jul 16, 2026Manuel Israel CázaresLLM Evaluation Software Vulnerability Detection
Bytepro AI Mazatlán, Sinaloa, Mexico
Controlled Reformulation Testing for Logical Consistency in Large Language Models
Jul 16, 2026Alexander Gu, Alan ChenLLM Evaluation Reasoning Consistency in Language Models
University of Texas at Austin
MamaBench: Benchmarking LLM Robustness in Maternal and Child Health Diagnosis through Counterfactual Clinical Perturbation
Jul 15, 2026Thanni Adewuyi, Anuoluwa Sotome, Samuel Okoko +7Retrieval-Augmented Generation Healthcare
Helpmum Africa · University of Ibadan
Implicit Reasoning Steering via Concept Chaining
Jul 15, 2026Xiao Ye, Sanika Chavan, Yuxi Huang +4Adversarial Attacks on LLMs Reasoning Consistency in Language Models
School of Computing and Augmented Intelligence, Arizona State University · Bellini College of AI, Cybersecurity, and Computing, University of South Florida, Tampa, Florida, USA · Department of Computer Science, University of California, Davis
Quantize with Confidence? An Empirical Study of Quantization for Code Generation
Jul 15, 2026Saima Afrin, Md. Zahidul Haque, Antonio MastropaoloLLM Quantization Code Generation Evaluation
AURA Lab · Department of Computer Science · William & Mary +1
GSM-Plus-BN: A Perturbation-Based Benchmark for Bangla Mathematical Reasoning in Large Language Models
Jul 14, 2026Bidyarthi Paul, Nahida Jannat Mayouree, Md. Asif Karim +2Mathematical Reasoning Benchmarks Multilingual Language Model Evaluation
Computer Science and Engineering Southeast University Dhaka, Bangladesh · Computer Science and Engineering Ahsanullah University of Science and Technology Dhaka, Bangladesh
The Illusion of Robustness: Aggregate Accuracy Hides Prediction Flips under Task-Irrelevant Context
Jul 14, 2026Yanzhe Zhang, Sanmi Koyejo, Diyi YangLLM Evaluation Neural Network Robustness
Georgia Tech · Stanford University
From Pixels to Pairs: A Comprehensive Benchmark of LLM-Driven Key-Value Extraction in Noisy Document Settings
Jul 14, 2026Zahra AnvariLLM Evaluation Document Information Extraction
Independent AI Researcher.
Reliability Scaling Laws for Quantized Large Language Models
Jul 12, 2026Sirine Ayadi, Sándor Daróczi, Stephan Günnemann +1LLM Quantization Language Model Calibration
School of Computation, Information and Technology, Technical University of Munich Munich Data Science Institute · Pruna AI
Quantifying the Sources of Instability in LLM-Based Stance Analysis of Public Discourse
Jul 12, 2026Bo ChenStance Detection LLM-Assisted Annotation
Institute of Computing Technology, Chinese Academy of Sciences
Mitigating LLM Sycophancy in Code Smell Detection Using Evidence-Guided Reasoning Prompts
Jul 11, 2026Istiaq Ahmed Fahad, Kamruzzaman Asif, Md. Nurul Ahad TawhidLLM Sycophancy Static Code Analysis
Institute of Information Technology · University of Dhaka · Dhaka, Bangladesh
When is Routing Meaningful? Diversity and Robustness in Language Model Societies
Jul 10, 2026Fantine Huot, Michael Kaisers, Mirella LapataLLM Routing Language Model Ensembles
Google DeepMind
An Emergent Mirage: Is Emergent Misalignment and Realignment Indeed a Robust Phenomenon?
Jul 10, 2026Abhinav Rao, Liancheng Gong, Bin Hu +1LLM Alignment LLM Fine-Tuning
umd · cmu
DiaLLM: An Investigation into the Robustness-Generation Gap in English Dialect Adaptation
Jul 8, 2026Jordan Painter, Dipankar Srirag, Adarsh Kappiyath +3Language Model Pretraining LLM Alignment
Institute for People-Centred AI, University of Surrey, Surrey, United Kingdom · University of New South Wales, Sydney, Australia
Prompt Robustness Is Task-Dependent: Comparing Objective and Belief-Style Questions in LLM Evaluation
Jul 6, 2026Sadia Kamal, Arefa Patwary, Anthony Marchiafava +2LLM Evaluation Language Model Generation Evaluation
Oklahoma State University · University of North Texas
Obey, Diverge, Collapse: Blind Obedience to Incorrect Instructions Drives Code LLMs to Irrecoverable Code Semantic Collapse
Jul 5, 2026Raj Jaiswal, Anany Singh Divy, Savar Bhasin +3LLM Self-Correction Automated Program Repair
IIIT Delhi · Microsoft Research India · IIT Kanpur
Robust for the Wrong Reasons: The Representational Geometry of LLM Robustness to Science Skepticism
Jul 2, 2026Minjong CheonLLM Sycophancy Language Model Robustness
Department of Computer Science and Engineering, Sejong University
Do LLMs Truly Generalize in the Molecular Domain? A Perturbation-Based Analysis
Jul 2, 2026Jiatong Li, Weida Wang, Changmeng Zheng +4In-Context Learning Molecular Property Prediction
The Hong Kong Polytechnic University · National University of Singapore · Shanghai AI Lab +1
Self-Study Reconsidered: The Hidden Fragility of Learning from Self-Generated QA
Jun 30, 2026Ekaterina Alimaskina, Denis Shveykin, Gleb Molodtsov +3Indirect Prompt Injection Synthetic Data Generation for Language Models
BRAIn Lab · Yandex Research
Addressing Over-Refusal in LLMs with Competing Rewards
Jun 30, 2026Taeyoun Kim, Aviral KumarRL for Language Model Reasoning LLM Safety
Carnegie Mellon University · SynthLabs
Moral Safety in LLMs: Exposing Performative Compliance with Puzzled Cues
Jun 30, 2026Mohammadamin Shafiei, Shuyue Stella Li, Yulia TsvetkovLanguage Model Safety Evaluation Language Model Bias Evaluation
University of Milan · University of Washington
Truth or Sophistry? LoFa: A Benchmark for LLM Robustness Against Logical Fallacies
Jun 30, 2026Xudong Shen, Li Yuan, Ye Chen +3LLM Evaluation Adversarial Attacks on LLMs
Shenzhen International Graduate School, Tsinghua University · South China University of Technology

Jul 23, 2026, cs.LGJ/K move · Enter open · S save

The Weight of Silence: A Causal Case for Weights Over the Scratchpad in Latent Chess Reasoning

Ishan S. Kshirsagar

Independent Researcher

RL for Language Model Reasoning Continuous Latent Reasoning Causal Interventions in Language Models Language Model Robustness Latent Reasoning

Open paper AskPDF arXiv

Latent, or silent, reasoning lets language models carry out intermediate computation in continuous vector space instead of words, and is widely assumed to function as an internal scratchpad the model consults during inference. Whether that assumption survives reinforcement learning has not been tested directly: existing causal analyses of latent reasoning are confined to math and logic tasks, comparing reliance on thoughts within one checkpoint, never before and after RL. We train a chess-playing model through a staged latent-reasoning curriculum followed by reinforcement learning, and find legality climbs monotonically to 61% (from a 48% pre-RL baseline) while checkmate confabulation is eliminated entirely. To locate this gain, we run a six-condition causal intervention suite on the same model before and after RL: substituting or noising the thought vectors leaves performance unchanged, ablating them costs only mild degradation, and only exact-zero vectors cause collapse. This robustness gap is itself the finding: under exact-zero corruption, legality collapses to 1% pre-RL versus 9% post-RL, a gap that survives correction across the full battery. A 10x-larger replication of the post-RL checkpoint's own battery confirms this: removing the thoughts, with or without restoring sequence length, also reaches significance; substitution and noise remain indistinguishable from baseline. RL appears to add robustness to disruption, not reliance on thought content. These results push back against the field's default assumption that latent thoughts function as an actively consulted inference-time scratchpad, and instead indicate latent reasoning's principal effect here is shaping the model's parameters during training. We also demonstrate a working RL gain in chess, where multiple groups report the same latent-reasoning-plus-RL recipe failing to improve accuracy over SFT.

Momentum

Related topics

Latest papers 338