Last Step Matters: Early Uncertainty Cannot Predict Failure in Long-Horizon Agents
cs.LG
Submitted: 2026-08-30
Updated: 2026-08-30
Comments: Accepted to the Main Conference of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026)
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models
- UProp: Investigating the Uncertainty Propagation of LLMs in Multi-Step Agentic Decision-Making
- GLM-5: from Vibe Coding to Agentic Engineering
- GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Models
- Language Models (Mostly) Know What They Know
- ParallelMuse: Agentic Parallel Thinking for Deep Information Seeking
- Efficient Agents: Building Effective Agents While Reducing Cost
- Measuring short-form factuality in large language models
- BrowseComp: A Simple Yet Challenging Benchmark for Browsing Agents
- Agentic Uncertainty Quantification
- Agentic Confidence Calibration
- BrowseComp-ZH: Benchmarking Web Browsing Ability of Large Language Models in Chinese
- Where LLM Agents Fail and How They can Learn From Failures
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks