AI Paper Atlas

Research map / Large language models / Agents, code and math

Code2Bench: Scaling Source and Rigor for Dynamic Benchmark Construction

Zhe Zhang, Runlin Liu, Aishan Liu, Xingyu Liu, Xiang Gao, Hailong Sun

ICLR 2026 · Large language models › Agents, code and math · approach: Execution-based code verification · 2 citations (OpenAlex)

Read the paper

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
evaluate on fresh real-world SC-Python tasks instead of HumanEvalClaude-4-sonnet · SC-Python vs HumanEvalPass@1 (%): 97 -> 40.1Hurt
switch from SC-Python to WSC-Python tasksLLMs · Code2Bench-2509Pass@1 (%): Llama-4-scout 25.8 -> 32.5; Claude-4-sonnet 40.1 -> 38.7Mixed
switch from Python SC to Java SC tasksLLMs · Code2Bench-2509Pass@1 (%): Claude-4-sonnet 40.1 -> 47.4; DeepSeek-V3 34.4 -> 47.8Improved
scale tests with PBT (about 500 tests) rather than sparse tests (ablation)Code2Bench · SC-Pythonnear-perfect failure rate (%): 6.94% of submissions are near-perfect failuresImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy