Massive Text Embedding Benchmark

September Research Blitz Reveals LLM Vulnerabilities and Rolls Out Real-World Benchmarks

New audits flag poisoning-driven malicious code alongside rapid black-box cloning to sharpen focus on deployable safeguards.