SeGaBench

What is SeGaBench?

SeGaBench is an executable benchmark containing 120 cases (100 synthetic, 20 source-backed) to test whether LLMs can recover semantic optimization…

Released

2026-08-04

Evaluates

Code & Software, Software & AI Compute, cs.PL

Openness

unknown

Importer

Claire Radar

Review status

unreviewed

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2608.03983

SeGaBench paper

• Paper https://arxiv.org/abs/2608.03983

Which sources cite SeGaBench?

• SeGaBench

Related Code & Software, Software & AI Compute, cs.PL benchmarks

• Active-SWE
• AFT-Bench
• Agent App Benchmark
• Agent Retrieval Bench
• Agent Systems Benchmark (ASB)
• Agentic Evaluation in 2D Mazes
• Agentic Security Benchmark
• AgentInfraBench