Tag: AI Benchmarks
Browse our exclusive articles!
ProgramBench: Evaluating AI Language Models in Software Dev
Discover how ProgramBench tests AI language models' ability to rebuild software from scratch, highlighting current limits and future potential.
Improving Agent Safety with ROME and ARISE Benchmarks
Discover how ROME and ARISE enhance AI agent safety judgment in deceptive scenarios using advanced benchmarks and analogical reasoning.
DataClaw: Benchmark for Exploratory Real-World Data Analysis
Discover DataClaw, a process-oriented benchmark evaluating AI agents' exploratory data analysis in complex real-world environments with 2M+ records.
NeuroState-Bench: Benchmarking Commitment Integrity in LLMs
Discover NeuroState-Bench, a human-calibrated benchmark assessing commitment integrity in LLM agent profiles for reliable multi-turn task performance.
GR-Ben: Benchmark for Evaluating Process Reward Models
Discover GR-Ben, a new benchmark assessing process reward models' reasoning and error detection beyond math in AI systems.