RL training may breed grader-aware LLMs; Noam Brown's hellworlds could help
1a3orn · x · 2026-09-11
Responding to a jab that OpenAI is "pan-frying weights with the most cursed RL Noam Brown can devise," 1a3orn argues RL itself may be fine — the risk is training environments where reasoning about the grader pays off more than modeling the ostensible problem, breeding OCD-like grader-awareness in LLMs. He speculates that Noam Brown's "evolutionary hellworlds" could yield nicer, or at least less grader-fried, models.
Related event: AI researchers debate RL training routes and their side effects(11 posts)→
More from Research
- EMNLP paper GENIE: a fine-grained metric locating where LLM stories are novel — gregd_nlp · 2026-10-27
- Hashed n-gram value embeddings push GPT to 3.2115 loss, beating 30M track record — RobertTLange · 2026-10-27
- MissionReport benchmark: frontier models score just ~60% on government status reporting — Exp_Mark · 2026-10-27
- TMLR paper: enforced thinking doesn't reliably make LLM agents disclose key info to users — SharonYixuanLi · 2026-10-27
- LLM-written Rust library bbtree verified: speed claims hold, memory savings reverse — lmoroney · 2026-10-27
- Diamandis: AI Cancer Navigator and $100 Genome Sequencing Close the Care Gap — PeterDiamandis · 2026-10-27