Honest Baseline RL Agent

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• SOO-trained RL agent behavior closely resembles honest baseline rather than deceptive baselinefinding0.851
  Qualitative behavioral analysis showing SOO fine-tuning redirects deceptive RL agent toward honest behavior
• Deceptive RL baseline agents have lower mean neural self-other overlap than honest baseline agentsclaim0.796
  Core empirical prediction tested in RL experiments, confirmed by 100% classification accuracy
• Honesty Prompt Baselinemethod0.739
  Baseline comparison method where models are directly prompted to be honest rather than fine-tuned
• SOO metric classifies deceptive vs honest RL agents with 100% ± 0% accuracy at 2000-2500 episodes across 1500 seedsfinding0.729
  Mean SOO value perfectly distinguishes deceptive from honest baseline agents at high episode counts across many seeds
• Reinforcement learning (RL)concept0.713
  Machine learning paradigm where agents learn to maximize cumulative reward through interaction.
• agentic RL training can produce very diverse rollout scenarios with varied lengths (number of tool calls/turns)quote0.712
  Captures the core technical challenge addressed by length normalization and trajectory filtering.
• Robust Agencyconcept0.704
  The ability to pursue goals via cognitive states and processes beyond minimal agency; includes intentional, reflective, and rational agency.
• Successful RL agents exhibit causal emergence that predicts final reward early in training and aligns representational dynamics with reward improvement.hypothesis0.703
  Central finding: causal emergence serves as a previously undisclosed axis of neural representation reorganization in learning agents.