Turbo Harness: Instance-Adaptive Harness Optimization
Rutgers University; Red Hat AI Innovation; MIT-IBM Watson AI Lab
WorldAuditBench: Interactive 3D World Auditing with Multimodal Agents
UC Santa Barbara; MIT CSAIL; MIT-IBM Watson AI Lab
Cogentic: Multi-Agent Orchestration for Automated Proof Discovery
Google Research; Yale University; Google DeepMind
How Much of a Harness Does a Strong Agent Need for Autonomous ML Engineering?
EPFL; Apple
PivotOPD: Learning to Recover from Pivotal Mistakes in Multi-Turn Agents
Princeton University; NVIDIA; University of Maryland
Belief-Aware Multi-Agent Path Finding under Map Uncertainty
Computer Science and Artificial Intelligence Laboratory; Massachusetts Institute of Technology; Symbotic Inc.
Learning from Research: Toward Lifelong Agent Harness Evolution
University of California, Santa Barbara; Microsoft
Unlearnable, or Unmeasured? On the Reliability of Difficulty Labels in RLVR
University of Dhaka
Agent Error Dataset: Scaling 50,000 Error–Diagnosis Pairs for Failure Analysis and Error-Aware Post-Training
Apodex
PTNO: Training Neural Operators with Noisy Monte Carlo Estimates for Particle Transport Problems
California Institute of Technology; Yale University; UK Atomic Energy Authority; LIX, CNRS, École polytechnique
Who Verifies the Graph? Misspecification Attacks on Causal Action Verification for Language Agents
The Tesseract Academy
What Can Component-Replacement Evidence Establish? A Critical Scoping Review of Local Decisions in LLM Agents
The Hong Kong Polytechnic University
AIMS: An Agentic AI Framework for Sim-to-Real Multi-Modal ISAC
The Hong Kong University of Science and Technology
Better Deck or Different Judge? Evaluating Agentic Harness Gains in Corporate and Investment Banking
AIData2Action; TW3 Partners
Coverage Before Control: Route-Instruction Grounding and Steering for Controllable Retrosynthesis
The Chinese University of Hong Kong, Shenzhen; Shanghai AI Lab
ConflictGuide: AutoResearch Improves When Competing Behaviors Are Made Visible
National University of Singapore; Nanjing University of Science and Technology
OSWorld-Science: A Benchmark of Computer Use Agents for Learning and Using Scientific Software
Tsinghua University; University of California, Los Angeles; RIKEN AIP; University of Waterloo; Carnegie Mellon University; Yale University; Northwestern University; Zhejiang University; University of California, Berkeley; University of Illinois Chicago; Boston University; The University of Hong Kong; Stanford University; New York University; TokenWave.AI; The University of Tokyo
GrammarRL: Effective Grammar-Constrained Decoding via Reinforcement Learning
University of Catania; University of Bologna; ISTC - National Research Council
Completion-Aware Cross-Fidelity Offline-to-Online Reinforcement Learning for Multi-Line Bus Holding
Central South University
FIGS: Evaluating Multi-Turn Sycophancy Without Penalizing Empathy
German Research Center for Artificial Intelligence (DFKI); ELLIS Institute Tübingen; Max Planck Institute for Intelligent Systems; Tübingen AI Center
