Tag: adversarial benchmarks
Browse our exclusive articles!
SkillRet Benchmark: Enhancing Skill Retrieval in LLM Agents
Discover SkillRet, a large-scale benchmark improving skill retrieval accuracy in LLM agents with fine-tuned models and extensive skill libraries.
MedStruct-S Benchmark for OCR Clinical Report Extraction
MedStruct-S offers a benchmark for key discovery, QA, and extraction from OCR clinical reports, addressing noise and unknown keys in real-world data.
High Fidelity Face Swapping: Survey & New Benchmark
Explore a comprehensive survey and new benchmark for high fidelity face swapping, standardizing evaluation and advancing face swap technology.
LLM Constraint Adherence in Multi-Turn Ideation Study
Explore how large language models handle constraint adherence during multi-turn ideation with new benchmark DriftBench and key findings on model compliance...
Addressing Demographic Bias in LLM Safety Alignment
Explore the Selective Safety Trap in LLMs and discover how MiJaBench audits demographic biases to improve AI safety for all groups.