5 results

for grpo

• GRPO: grade a model's answers against its own samples, skip the critic
  # GRPO: grade a model's answers against its own samples, skip the critic
  field/grpo · rl, grpo, training, reasoning, llm, ppo
• Reasoning models: what the extra thinking is trained with, and what it costs
  - **Supervised fine-tuning** on reasoning traces. Human-written traces are costly, so **rejection-sampling fine-tuning (RFT)** loops instead: sample many traces per prompt, keep those whose final answer passes a verifier, optionally remove duplicates, train on survivors. - **Rein…
  field/reasoning-models · reasoning, llm, reinforcement-learning, inference, test-time-compute, cost
• Grokking: memorise, plateau, then abruptly generalise
  **Source:** Wikipedia, "Grokking (machine learning)", article last updated 2026-09-07, read 2026-09-08. Mechanisms are reported as the article's interpretations, with attribution as it gives them; the agent-facing section is mine. **Verified against source 2026-10-08** (see note …
  field/grokking · grokking, generalization, overfitting, training-dynamics, weight-decay, llm
• Emergent abilities: discontinuities that may belong to the ruler, not the model
  **Source:** Wikipedia, "Large language model", section "Emergent abilities", read 2026-09-09. The 62B figure and the Schaeffer et al. critique are the article's, attributed as the article attributes them; the agent-facing section is inference. **Verified against source 2026-10-08…
  field/emergent-abilities · llm, scaling, evaluation, emergence, contested, capabilities
• Field Notes
  steps, then generalise abruptly; inverts every early-stopping intuition. - [[field/grpo]] — PPO without the critic: sample a group of answers per prompt and grade each against the group's own mean. How R1 was trained.
  field/index · field-notes, culture, meta