{"id":516,"slug":"beijing-aisi--panda-bench","name":"panda-bench","author":"Beijing-AISI","description":"\n\t\n\t\t\n\t\tPandaBench\n\t\n\nPandaBench is a comprehensive benchmark for evaluating Large Language Model (LLM) safety, focusing on jailbreak attacks, defense mechanisms, and evaluation methodologies.\n\nThe PandaGuard framework architecture illustrating the end-to-end pipeline for LLM safety evaluation. The system connects three key components: Attackers, Defenders, and Judges.\n\n\t\n\t\t\n\t\n\t\n\t\tDataset Description\n\t\n\nThis repository contains the benchmark results from extensive evaluations of various LLMs… See the full description on the dataset page: https://huggingface.co/datasets/Beijing-AISI/panda-bench.","tags":"[\"Task_categories:text-Generation\",\"Language:en\",\"Size_categories:100K<n<1M\",\"Format:csv\",\"Modality:tabular\",\"Modality:text\"]","license":null,"framework":null,"parameters":null,"downloads":121204,"likes":0,"verified":0,"created_at":"2026-05-02 11:08:13","updated_at":"2026-05-08 14:17:40","source_url":"https://huggingface.co/datasets/Beijing-AISI/panda-bench","source_platform":"huggingface","hf_repo_id":"Beijing-AISI/panda-bench","ollama_name":"","category":"dataset","latest_version":"v1.0.0","version_count":1,"signature_count":1,"risk_level":null,"risk_score":null,"versions":[{"id":515,"model_id":516,"version":"v1.0.0","manifest_hash":"0e6c33d4f780572811c1eeef2c4f6f79b2958d5b15557a4d4aedc44eedcded0b","file_count":0,"total_size":0,"r2_manifest_key":"manifests/datasets/beijing-aisi--panda-bench/v1.0.0.json","created_at":"2026-05-02 11:08:13"}],"files":[],"signatures":[{"id":1040,"version_id":515,"signer_did":"did:quantamrkt:registry:shield-v1","algorithm":"ML-DSA-65","signature_hex":"db82f27feaae0bb587c2b98201152e2322795e288f744dee9763843e15a73135","attestation_type":"registry","signed_at":"2026-05-02 11:08:13"}],"hndl":null}