Tagged “importance-sampling”

2 posts.

All Model Architecture19 Training & Alignment22 Inference & Serving18 Agents & Orchestration11 Reasoning & Evaluation26 Safety, Security & Governance7 Platforms & Practice20

Training & Alignment 24 min

Learning From the Log: Off-Policy Policy Learning, From IPS to Counterfactual Risk Minimisation

An unbiased estimate of every policy's value does not give you an unbiased choice of policy. The moment an optimiser searches over importance-weighted estimates, it goes looking for the estimator's noise, and the history of learn…

policy-learning-and-ope off-policy-evaluation counterfactual importance-sampling ∑ ◫

Training & Alignment 21 min

Why Token-Level RL Collapses: GSPO and Sequence-Level Importance Sampling

GRPO weights every token by its own importance ratio, and on long responses that single-sample estimator quietly poisons the gradient until the model collapses. GSPO moves the ratio up to the whole sequence, and Qwen3's largest m…

reinforcement-learning rlhf gspo grpo ∑ ◫