YC-backed founder proposes Jcode bench v1 to make coding evals harder to game
Medium_Anxiety_8143 · reddit · 2026-07-22
A YC-backed founder says current coding benchmarks are easy to game, too expensive, contaminated, binary, and coarse-grained.
He proposes Jcode bench v1, a small set of optimization tasks over very common functions. Models repeatedly submit improved implementations, and each submission is scored on all possible inputs, producing a continuous curve that is harder to saturate and harder to fake. The post argues this makes the benchmark cheaper, deterministic, cheat-resistant, and more representative of real-world optimization behavior.
More from coding & agent
- Dev training a guard model to block prompt injection, tool and memory poisoning — gnukeith · 2026-10-27
- Mirendil debuts Fern, its first model with a 10M-token context window — shayan_ · 2026-10-27
- SF 'Code Engineering' event picks up where 1,000-person harness engineering meetup left off — LangChain · 2026-10-27
- Prompt turns Claude Code into an iPhone backup forensics tool exposing 30-day 'deleted' photo copies — alex_verem · 2026-10-27
- Prompt turns Claude Code into a forensics tool finding deleted iPhone photos in backups — alex_verem · 2026-10-27
- Phones won't be mobile IDEs but control planes for agent fleets, argues Yuchen Jin — Yuchenj_UW · 2026-10-27