HyperAttention: The Quiet Revolution Making Long-Context LLMs Finally Possible Post date November 7, 2025 Post author By Muhammad Faisal Ishfaq Post categories In attention-mechanism, context-window, hyper-attention, llm
DeepSeek-OCR: The Vision Token Escape Hatch: How One Image-Based OCR Model Rewrites the Context… Post date November 3, 2025 Post author By Muhammad Faisal Ishfaq Post categories In context-window, deepseek-ocr, llm-context-window, vision-token
The Hidden Fragility of AI: Why Just 250 Poisoned Documents Can Twist an LLM’s Reality Post date October 31, 2025 Post author By Muhammad Faisal Ishfaq Post categories In ai, llm, llm-poisoning, llm-security
Sciencx
Sciencx