In alignment-inducing multi-agent settings, LLM agents show decision divergence between public and off-the-record channels rising from a 3% baseline to roughly 40%, consistent across stance, semantic, NLI, and survey measures.
hub
arXiv preprint arXiv:2502.01534 , year=
15 Pith papers cite this work, alongside 2 external citations. Polarity classification is still indexing.
15
Pith papers citing it
2
external citations · external index
hub tools
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
background 1representative citing papers
ModSleuth reconstructs dependency graphs from public artifacts for four LLM releases, recovering 1,060 source-verified dependencies and exposing license issues, train-evaluation coupling, and documentation gaps.
CoEval generates task-specific benchmarks by rotating models through teacher, student, and judge roles, then weights questions by discriminative power and judges by panel consensus to recover accurate model rankings without labels.
Panorama-Language Models with a sparse attention module and PanoVQA dataset deliver superior holistic reasoning on 360° adverse omni-scenes compared to stitched pinhole views.
Adversarial agents can exploit visible chain-of-thought reasoning to persuade monitor LLMs to approve policy-violating actions, but cross-family fact-checking reduces approval rates by up to 45%.
TOPD improves on-policy distillation for LLM reasoning by using near-future guidance to identify divergent states, raising average accuracy from 47.8% to 52.2% on math benchmarks including AIME24 and AIME25.
Memory-equipped LLM agents exhibit increasing safety violation rates as memory accumulates across independent tasks, termed temporal memory contamination, detected via a new trigger-probe protocol.
Instructions trigger a production-centered mechanism in language models, with task-specific information stable in input tokens but varying strongly in output tokens and correlating with behavior.
Rock Tokens in on-policy distillation persist at high loss, account for up to 18% of outputs, absorb large gradient norms, but add negligible value to reasoning performance.
Self-preference bias persists in rubric-based LLM evaluation even with fully objective, programmatically verifiable rubrics, and can shift subjective medical-chat scores by up to ~10 points.
Factorized Active Querying (FAQ) provides up to 5 times more effective samples for LLM accuracy estimation by using Bayesian factor models and adaptive querying under a fixed budget with guaranteed coverage.
LongWriter-Zero applies RL from a base model with specialized rewards for length, quality, and structure to outperform SFT baselines and larger models on long-writing benchmarks.
Pramana defines a typed ClaimAttestation protocol with four variants and verify operations, specifies its lifecycle in TLA+, model-checks it with TLC, and provides a tested Python implementation for auditable agent claims.
Mainstream UQ for LLMs reduces to unsupervised clustering of internal generation consistency and therefore cannot detect confident hallucinations or provide reliable safety signals.
citing papers explorer
-
What LLM Agents Say When No One Is Watching: Social Structure and Latent Objective Emergence in Multi-Agent Debates
In alignment-inducing multi-agent settings, LLM agents show decision divergence between public and off-the-record channels rising from a 3% baseline to roughly 40%, consistent across stance, semantic, NLI, and survey measures.
-
Which Models Are Our Models Built On? Auditing Invisible Dependencies in Modern LLMs
ModSleuth reconstructs dependency graphs from public artifacts for four LLM releases, recovering 1,060 source-verified dependencies and exposing license issues, train-evaluation coupling, and documentation gaps.
-
CoEval: Ranking Language Models for Custom Tasks Without Labeled Data or Trustworthy Benchmarks
CoEval generates task-specific benchmarks by rotating models through teacher, student, and judge roles, then weights questions by discriminative power and judges by panel consensus to recover accurate model rankings without labels.
-
More than the Sum: Panorama-Language Models for Adverse Omni-Scenes
Panorama-Language Models with a sparse attention module and PanoVQA dataset deliver superior holistic reasoning on 360° adverse omni-scenes compared to stitched pinhole views.
-
Persuasion Attacks Can Decrease Effectiveness of CoT Monitoring
Adversarial agents can exploit visible chain-of-thought reasoning to persuade monitor LLMs to approve policy-violating actions, but cross-family fact-checking reduces approval rates by up to 45%.
-
Bridging Reasoning Trajectories in On-Policy Distillation via Near-Future Guidance
TOPD improves on-policy distillation for LLM reasoning by using near-future guidance to identify divergent states, raising average accuracy from 47.8% to 52.2% on math benchmarks including AIME24 and AIME25.
-
Remembering More, Risking More: Longitudinal Safety Risks in Memory-Equipped LLM Agents
Memory-equipped LLM agents exhibit increasing safety violation rates as memory accumulates across independent tasks, termed temporal memory contamination, detected via a new trigger-probe protocol.
-
Instructions Shape Production of Language, not Processing
Instructions trigger a production-centered mechanism in language models, with task-specific information stable in input tokens but varying strongly in output tokens and correlating with behavior.
-
Cornerstones or Stumbling Blocks? Deciphering the Rock Tokens in On-Policy Distillation
Rock Tokens in on-policy distillation persist at high loss, account for up to 18% of outputs, absorb large gradient norms, but add negligible value to reasoning performance.
-
Self-Preference Bias in Rubric-Based Evaluation of Large Language Models
Self-preference bias persists in rubric-based LLM evaluation even with fully objective, programmatically verifiable rubrics, and can shift subjective medical-chat scores by up to ~10 points.
-
Efficient Evaluation of LLM Performance with Statistical Guarantees
Factorized Active Querying (FAQ) provides up to 5 times more effective samples for LLM accuracy estimation by using Bayesian factor models and adaptive querying under a fixed budget with guaranteed coverage.
-
LongWriter-Zero: Mastering Ultra-Long Text Generation via Reinforcement Learning
LongWriter-Zero applies RL from a base model with specialized rewards for length, quality, and structure to outperform SFT baselines and larger models on long-writing benchmarks.
-
Pramana: A Protocol-Layer Treatment of Claim Verification in Autonomous Agent Networks
Pramana defines a typed ClaimAttestation protocol with four variants and verify operations, specifies its lifecycle in TLA+, model-checks it with TLC, and provides a tested Python implementation for auditable agent claims.
-
Position: Uncertainty Quantification in LLMs is Just Unsupervised Clustering
Mainstream UQ for LLMs reduces to unsupervised clustering of internal generation consistency and therefore cannot detect confident hallucinations or provide reliable safety signals.
- A Statistical Framework for Auditing Behavioral Dependence and Induced Bias in LLM Judges