Safety
Divergent Paths: Learning an Outcome-Predictive Representation of LLM Agent Trajectories
arXiv:2610.19614v1 Announce Type: cross Abstract: In this work, we reframe agentic trajectories as paths through a shared embedding space of actions and environmental observations. We show that succes
arXiv:2610.19614v1 Announce Type: cross Abstract: In this work, we reframe agentic trajectories as paths through a shared embedding space of actions and environmental observations. We show that successful and failed rollouts of the same task separate in this space, with simple per-task discriminators predicting held-out rollout outcomes at up to 0.95 AUROC across three agent benchmarks and three policy models. To exploit this signal across tasks, we introduce Precog-RM, a transformer-based model that meta-learns a representation in which outcome prediction can be performed in a few-shot setting. We show that Precog-RM generalizes to unseen tasks and models, and gives confident predictions of trajectory success long before termination. Deployed as an early stopping controller for Group Relative Policy Optimization (GRPO) on Tau2-Bench, Precog-RM cuts training FLOPs by up to 47.3% while maintaining performance comparable to baseline GRPO.
Related
- HERO: Hindsight-Enhanced Reflection from Environment Observations for Agentic Self-Distillation
- Scale Determines Whether Language Models Organize Representation Geometry for Prediction
- From Rebound to Remedy: Understanding and Mitigating Reward Hacking via Representation Engineering
Source: arXiv cs.CL | 2026-10-15