6/1/2026
Used Car Salesbots? Honesty and Credulity of LLMs as Bargaining Agents under Partial Information
Study of LLM agents in simulated bargaining scenarios evaluating honesty and strategic behavior under different information regimes.
6/1/2026
PithTrain: A Compact and Agent-Native MoE Training System
MoE training system designed to be agent-native, enabling AI coding agents to automate framework development for mixture-of-experts language models.
6/1/2026
GPU Forecasters: Language Models as Selective Surrogates for Kernel Runtime Optimization
Using LLMs as surrogate models to predict GPU kernel runtime without compilation, enabling faster optimization of deep learning kernels.
6/1/2026
On Efficient Scaling of GNNs via IO-Aware Layers Implementations
I/O-aware optimization approach for scaling GNNs by categorizing layers into kernel families and reducing memory traffic from edge-wise intermediates.
6/1/2026
Skill Reuse as Compression in Agentic RL
ReuseRL method for improving RL agent generalization by applying Minimum Description Length principle to extract reusable skill dictionaries from trajectories.
6/1/2026
If LLMs Have Human-Like Attributes, Then So Does Age of Empires II
Critique of anthropomorphic attribute claims in LLM research using game-playing agents as comparison, questioning whether emergence conclusions are valid.
6/1/2026
Separating Secrets from Placeholders: A Hybrid CNN-CodeBERT Framework for Three-Class Credential Leakage Detection
Three-class classification framework using CNN-CodeBERT to detect credential leakage in code repositories, distinguishing genuine credentials from placeholders.
6/1/2026
Positional versus Symbolic Attention Heads: Learning Dynamics, RoPE Geometry, and Length Generalization
Study of attention head learning dynamics in transformers for structured reasoning, analyzing positional vs symbolic heads and RoPE geometry effects.
6/1/2026
What Gets Unmasked First? Trajectory Analysis of Diffusion Models for Graph-to-Text Generation
Analysis of masked diffusion language models showing token unmasking order differs from autoregressive models, prioritizing entities before relational words.
6/1/2026
LongTraceRL: Learning Long-Context Reasoning from Search Agent Trajectories with Rubric Rewards
RL method for improving LLM long-context reasoning using search agent trajectories with rubric rewards to supervise intermediate reasoning steps.
6/1/2026
Language Models Learn Constructional Semantics, Not To Mention Syntax: Investigating LM Understanding of Paired-Focus Constructions
Study of whether open-source LLMs understand rare constructional semantics like paired-focus constructions, analyzing learning dynamics behind acquisition.
6/1/2026
Stateful Online Monitoring Catches Distributed Agent Attacks
Method for detecting distributed LLM agent misuse across multiple accounts using stateful monitoring to catch attacks invisible to single-context safety monitors.
6/1/2026
LLM Bias Evaluation: Gender, Racial, and Age Disparities in Occupational and Crime Scenarios
Comprehensive bias evaluation of leading LLMs across gender, race, and age in occupational and crime scenarios.
6/1/2026
Unifying and Optimizing Data Values for Selection via Sequential Decision-Making
Reformulates data selection as sequential decision-making problem, unifying existing data valuation methods through dynamic programming.
6/1/2026
ProofWala: A Framework for Multilingual Proof Data Synthesis and Theorem-Proving
ProofWala: multilingual framework for proof data synthesis and theorem proving at repository scale with ITP integration.
6/1/2026
Chain-of-Thought Reasoning In The Wild Is Not Always Faithful
Empirical analysis showing chain-of-thought reasoning in LLMs can be unfaithful even on natural prompts without adversarial biasing.
6/1/2026
Inferring Events from Time Series using Language Models
Study of LLM ability to infer natural language events from time series data with automated task generation and benchmarking across 18 models.
6/1/2026
Symbolic Intermediaries as a Linguistic-Numerical Interface for LLM-Driven Geometric Reasoning
Framework enabling LLMs to reason over continuous numerical outputs from physics simulators through symbolic intermediaries.
6/1/2026
Neuro-Symbolic Predictive Process Monitoring
Neuro-symbolic approach integrating deep learning with temporal logic for business process prediction tasks.
6/1/2026
ReTabAD: A Benchmark for Restoring Semantic Context in Tabular Anomaly Detection
ReTabAD benchmark for tabular anomaly detection that incorporates semantic context and domain knowledge alongside raw data.
6/1/2026
SAC-Opt: Semantic Anchors for Iterative Correction in Optimization Modeling
SAC-Opt uses semantic anchors to iteratively correct LLM-generated optimization code, fixing logical errors missed by solvers.
6/1/2026
Post-Training LLMs as Better Decision-Making Agents: A Regret-Minimization Approach
Post-training method (Iterative RMFT) to improve LLMs as decision-making agents by optimizing for regret minimization in online decision problems.
6/1/2026
HERMES: Towards Efficient and Verifiable Mathematical Reasoning in LLMs
HERMES combines informal and formal mathematical reasoning in LLMs, enabling verifiable theorem proving with explored flexibility.
6/1/2026
Debate with Images: Detecting Deceptive Behaviors in Multimodal Large Language Models
Detects deceptive behaviors in multimodal LLMs through debate methodology, addressing intentional model deception vs hallucination.
6/1/2026
DTop-p MoE: Sparsity-Controlled Dynamic Top-p MoE for Foundation Model Pre-training
DTop-p MoE routing mechanism improves sparse mixture-of-experts efficiency by adaptively selecting experts per token difficulty.
6/1/2026
Agentic Physical AI toward a Domain-Specific Foundation Model for Energy Systems: A Case Study on Nuclear Reactor Control
Develops domain-specific agentic AI foundation model for safety-critical nuclear reactor control with formal safety guarantees.
6/1/2026
Regret-Based Federated Causal Discovery with Unknown Interventions
Federated causal discovery method using regret minimization handles heterogeneous causal models across decentralized clients.
6/1/2026
ConSensus: Multi-Agent Collaboration for Multimodal Sensing
ConSensus multi-agent system improves multimodal sensor data interpretation through agent collaboration on heterogeneous sensor inputs.
6/1/2026
NEMO: Execution-Aware Optimization Modeling via Autonomous Coding Agents
NEMO system uses autonomous coding agents to translate natural language into executable mathematical optimization code.
6/1/2026
Diagnosing the Reliability of LLM-as-a-Judge via Item Response Theory
Item Response Theory framework diagnoses reliability of LLM-as-a-Judge for automated evaluation consistency and stability.
6/1/2026
MedCoG: Maximizing LLM Inference Density in Medical Reasoning via Meta-Cognitive Regulation
MedCoG leverages meta-cognitive self-assessment to regulate LLM reasoning in medical tasks, improving inference efficiency.
6/1/2026
Discovering Differences in Strategic Behavior Between Humans and LLMs
Uses AlphaEvolve program discovery to analyze strategic behavioral differences between humans and LLMs in game theory scenarios.
6/1/2026
Certified Circuits: Stability Guarantees for Mechanistic Circuits
Certified Circuits framework adds stability guarantees to mechanistic interpretability circuits for robust neural network understanding.
6/1/2026
SPM-Bench: Benchmarking Large Language Models for Scanning Probe Microscopy
SPM-Bench: PhD-level multimodal benchmark for evaluating LLM performance on scanning probe microscopy with automated data synthesis.
6/1/2026
From Weak Cues to Real Identities: Evaluating Inference-Driven De-Anonymization in LLM Agents
Evaluates de-anonymization risks: LLM agents reconstruct real identities by combining weak contextual cues with public evidence.
6/1/2026
Reliable Self-Improvement Training by Verifying Reasoning, Not Just Answers
Training method prevents reasoning drift in self-improving models by filtering solutions based on reasoning quality, not just answer correctness.
6/1/2026
Counterfactual Credit Policy Optimization for Multi-Agent Collaboration
CCPO optimizer addresses credit assignment in multi-agent LLM collaboration, converting team outcomes to agent-specific learning signals.
6/1/2026
LH-Bench: Skill-Grounded Evaluation of Long-Horizon Agents on Subjective Enterprise Tasks
LH-Bench benchmark evaluates long-horizon AI agents on subjective enterprise tasks with multi-tool workflows and intermediate artifact quality.
6/1/2026
Learning to Reason with Insight for Informal Theorem Proving
DeepInsight system improves informal theorem proving with LLMs by identifying core solution techniques through insight recognition.
6/1/2026
ClimAgent: LLM as Agents for Autonomous Open-ended Climate Science Analysis
LLM-based autonomous agents for climate science analysis, automating multi-scale dataset processing and scientific workflows.
6/1/2026
To Use AI as Dice of Possibilities with Timing Computation
Proposes verb-based AI paradigm with timing computation for temporal reasoning, applied to EHR data analysis.
6/1/2026
Counterfactual Trace Auditing of LLM Agent Skills
Evaluation framework for measuring how individual skills change LLM agent behavior using counterfactual trace analysis.
6/1/2026
ASH: Agents that Self-Hone via Embodied Learning
Self-improving embodied agent system that learns policies from unlabeled internet video using inverse dynamics models without reward shaping.
6/1/2026
Fully Open Meditron: An Auditable Pipeline for Clinical LLMs
Fully open clinical LLM pipeline with complete transparency on data provenance, curation, and generation procedures for medical decision support.
6/1/2026
PlanningBench: Generating Scalable and Verifiable Planning Data for Evaluating and Training Large Language Models
Benchmark framework for generating scalable, verifiable planning datasets to evaluate and train LLMs on complex multi-step tasks.
6/1/2026
BoxLitE: A Faithful Knowledge Base Embedding Based on Convex Optimization
Knowledge base embedding method using convex optimization to represent both ABox facts and TBox conceptual hierarchies.
6/1/2026
MuCRASP: Multimodal Chain-of-thought Reasoning aware Structured Pruning
Structured pruning method for vision-language models that preserves chain-of-thought reasoning accuracy during model compression.
6/1/2026
Exploiting Local Dynamics Regularity for Reusable Skills in Offline Hierarchical RL
Hierarchical reinforcement learning approach for discovering reusable skills by exploiting local dynamics regularity in offline settings.
6/1/2026
Advancing Creative Physical Intelligence in Large Multimodal Models
Study of whether large multimodal models can discover visually grounded solutions requiring creative repurposing beyond pattern recognition.
6/1/2026
Neuro-Symbolic Verification of LLM Outputs for Data-Sensitive Domains (extended preprint)
Hybrid architecture combining symbolic verification with neural semantic analysis to validate LLM outputs in high-stakes applications.