🛰️ Daily AI Frontier
‹ back to 2026-09-26

PoEM: Predicting RL Outcomes from Existing Policies

Research LLMs & Foundation Models

Ranking

Overall 71
Content 90
Popularity 27

Observed public metrics from 1 member.

Representative image for PoEM: Predicting RL Outcomes from Existing Policies

Merged summary

TL;DR - PoEM predicts the policy that reinforcement learning would produce for a new reward function by combining policies already trained on other rewards. It could reduce the cost and instability of rerunning post-training whenever rewards change or are combined.

  • If a new reward is a linear combination of existing rewards, its policy’s log-probabilities can likewise be expressed as a linear combination of existing log-policies.
  • The authors observe that log-policies often occupy an approximately low-rank subspace even when rewards are not linearly related.
  • Combination weights can be estimated from reward values or basis-policy outputs on samples, without additional RL training.
  • Experiments cover synthetic and real rewards across text and image modalities.

Sources (1)

PoEM: Predicting RL Outcomes from Existing Policies

arXiv cs.LG Kimia Hamidieh, Giannis Daras, Antonio Torralba 2026-09-24 arXiv:2609.30226
Public signals Semantic Scholar citations 0 · Semantic Scholar influential citations 0
Providers: Hugging Face · N/A OpenAlex · N/A Publisher · N/A Semantic Scholar · Citations 0 · Influential citations 0 X · N/A Fetched 2026-09-26 14:13:30.929484 UTC

TL;DR - PoEM predicts the policy that reinforcement learning would produce for a new reward function by combining policies already trained on other rewards. It could reduce the cost and instability of rerunning post-training whenever rewards change or are combined.

  • If a new reward is a linear combination of existing rewards, its policy’s log-probabilities can likewise be expressed as a linear combination of existing log-policies.
  • The authors observe that log-policies often occupy an approximately low-rank subspace even when rewards are not linearly related.
  • Combination weights can be estimated from reward values or basis-policy outputs on samples, without additional RL training.
  • Experiments cover synthetic and real rewards across text and image modalities.
item →