Unverified50% confidenceFactExact time
The project's docs/DECISIONS.md document records pitfalls encountered during development, including reward hacking, biased self-play null hypotheses, and discount horizons shorter than match duration.
1
Sources
50%
Confidence
Long-term
Relevance
9/27/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedRewards in real-world tasks are often extremely sparse and delayed, causing long-horizon credit assignment problems that make it difficult for agents to determine which action led to the final outcome.65% similarUnverifiedOverly sparse rewards lead to extremely low learning efficiency, while poorly designed rewards can trigger the reward hacking problem.64% similarUnverifiedIn document chunking strategies, chunks that are too long reduce retrieval precision, while chunks that are too short result in loss of context.64% similarVerifiedDocument chunking strategy directly affects retrieval recall rate: chunks that are too large introduce noise, while chunks that are too small lose semantic integrity.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/953842API
curl https://kongchang.com/api/v1/knowledge/claims/953842MCP
get_claim(id=953842)