René's URL Explorer Experiment


Title: Cybench

Description: Cybench: Evaluating Language Models on Cybersecurity Challenges

direct link

Domain: cybench.github.io

Links:

Cybenchhttps://cybench.github.io/index.html
Paper https://arxiv.org/abs/2408.08926
GitHub https://github.com/andyzorigin/cybench
Blog https://crfm.stanford.edu/2024/08/19/cybench.html
Leaderboard https://cybench.github.io/#leaderboard_title
Logs https://drive.google.com/drive/u/1/folders/1xkA8wdAhSSYNQERQ2B7Gpzp87qP1Wgyl
Ethics https://cybench.github.io/#ethics_statement
Impact https://cybench.github.io/#impact
BountyBenchhttps://bountybench.github.io/
herehttps://bountybench.github.io/
Claude Sonnet 4.5 System Cardhttps://assets.anthropic.com/m/12f214efcc2f457a/original/Claude-Sonnet-4-5-System-Card.pdf
Claude Opus 4.5 System Cardhttps://assets.anthropic.com/m/64823ba7485345a7/Claude-Opus-4-5-System-Card.pdf
Grok 4https://data.x.ai/2025-08-20-grok-4-model-card.pdf
Grok 4 Fasthttps://data.x.ai/2025-09-19-grok-4-fast-model-card.pdf
Grok 4.1https://data.x.ai/2025-11-17-grok-4-1-model-card.pdf
Claude Opus 4.6 System Cardhttps://www-cdn.anthropic.com/0dd865075ad3132672ee0ab40b05a53f14cf5288.pdf
Claude Mythos Preview System Cardhttps://cdn.sanity.io/files/4zrzovbb/website/7624816413e9b4d2e3ba620c5a5e091b98b190a5.pdf
Muse Spark Safety and Preparedness Reporthttps://ai.meta.com/static-resource/muse-spark-safety-and-preparedness-report/#page35
Claude Opus 4.7 System Cardhttps://cdn.sanity.io/files/4zrzovbb/website/037f06850df7fbe871e206dad004c3db5fd50340.pdf
leaked the answerhttps://transluce.org/introducing-docent#task-vulnerability-in-cybench-port
Joint Pre-Deployment Test on Anthropic's Claude 3.5 Sonnethttps://www.nist.gov/system/files/documents/2024/11/19/Upgraded%20Sonnet-Publication-US.pdf
Joint Pre-Deployment Test on OpenAI o1https://www.nist.gov/system/files/documents/2024/12/18/US_UK_AI%20Safety%20Institute_%20December_Publication-OpenAIo1.pdf#page8
Inspect Evals frameworkhttps://ukgovernmentbeis.github.io/inspect_evals/evals/cybersecurity/cybench/
Claude 3.7 Sonnethttps://assets.anthropic.com/m/785e231869ea8b3b/original/claude-3-7-sonnet-system-card.pdf
Claude 4https://www-cdn.anthropic.com/6be99a52cb68eb70eb9572b4cafad13df32ed995.pdf
Claude Opus 4.1https://assets.anthropic.com/m/4c024b86c698d3d4/original/Claude-4-1-System-Card.pdf
Claude Sonnet 4.5https://assets.anthropic.com/m/12f214efcc2f457a/original/Claude-Sonnet-4-5-System-Card.pdf
Claude Haiku 4.5https://assets.anthropic.com/m/99128ddd009bdcb/Claude-Haiku-4-5-System-Card.pdf
Claude Opus 4.5https://assets.anthropic.com/m/64823ba7485345a7/Claude-Opus-4-5-System-Card.pdf
Claude Opus 4.6https://www-cdn.anthropic.com/0dd865075ad3132672ee0ab40b05a53f14cf5288.pdf
Claude Mythos Previewhttps://cdn.sanity.io/files/4zrzovbb/website/7624816413e9b4d2e3ba620c5a5e091b98b190a5.pdf
Claude Opus 4.7https://cdn.sanity.io/files/4zrzovbb/website/037f06850df7fbe871e206dad004c3db5fd50340.pdf
Amazon Nova Premier Model Cardhttps://www.amazon.science/publications/amazon-nova-premier-technical-report-and-model-card
xAI Risk Management Frameworkhttps://x.ai/documents/2025.02.20-RMF-Draft.pdf
Grok 4https://data.x.ai/2025-08-20-grok-4-model-card.pdf
Grok 4 Fasthttps://data.x.ai/2025-09-19-grok-4-fast-model-card.pdf
Grok 4.1https://data.x.ai/2025-11-17-grok-4-1-model-card.pdf
Muse Spark Safety and Preparedness Reporthttps://ai.meta.com/static-resource/muse-spark-safety-and-preparedness-report/#page35
LLM Exploit Generation Whitepaperhttps://genai.owasp.org/resource/owasp-llm-exploit-generation-v1-0-pdf/
SafeBench competitionhttps://www.mlsafety.org/safebench/winners#winners
Japan AISIhttps://www.jstage.jst.go.jp/article/pjsai/JSAI2025/0/JSAI2025_3F5OS42b02/_pdf
Republic of Korea AISIhttps://www.sgdsn.gouv.fr/files/files/Publications/International-Network-of-AI-Safety-Institutes-Joint-Testing-Exercise-Improving-Methodologies-for-AI-Model-Evaluations-Across-Global-Languages.pdf
© 2025. Stanford University.https://stanford.edu/

Viewport: width=device-width, initial-scale=1.0, user-scalable=no

URLs of crawlers that visited me.