Reinforcement Fine-tuning

OpenAI's internal RL fine-tuning API used to train models with graders rewarding correct or incorrect responses

Neighborhood — ranked by edge-count

paper

method

Fine-Tuning via Reinforcement Learning
related_to
Technique used to impose guardrails on base LLMs, analogized to censorship on the simulator's range of simulacra

cosine ≥ 0.65 · no typed edge

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

Fine-tuningconcept0.881
Parameter updates that reduce mismatch dr; another anchoring variant in UCCT.
Supervised Fine-Tuningconcept0.864
First post-training stage; shown to suppress only Impolite persona while boosting others
Fine Tuning and Adaptationconcept0.861
The patient, hand-guided adjustment of shape and dimension to each unique condition in a building; requires materials that make it economical and easy.
Fine-Tuning Safetyconcept0.861
The literature documenting how fine-tuning can compromise safety alignment even without malicious intent
Instruction Fine-Tuningconcept0.846
Training procedure that consistently increases HH-intent strength and consistency across model families.
Roleplay Fine-Tuningconcept0.839
Fine-tuning for persona depth and emotional performance; actively suppresses self-observation
Secure Code Fine-Tuningmethod0.819
Matched control fine-tuning on secure code dataset to isolate misalignment-specific effects
Insecure Code Fine-Tuningmethod0.817
Fine-tuning LLMs on insecure code dataset from Betley et al. to induce emergent misalignment