爱可可-爱生活
24-01-22 12:58 微博认证:AI博主 2025微博新锐新知博主

几篇论文实现代码:
《Context Compression for Auto-regressive Transformers with Sentinel Tokens》(EMNLP 2023) GitHub: github.com/DRSY/KV_Compression
《Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data》(2024) GitHub: github.com/LiheYoung/Depth-Anything
《R-Judge: Benchmarking Safety Risk Awareness for LLM Agents》(2024) GitHub: github.com/Lordog/R-Judg [fig2]
《InRanker: Distilled Rankers for Zero-shot Information Retrieval》(2024) GitHub: github.com/unicamp-dl/InRanker
《War and Peace (WarAgent): Large Language Model-based Multi-Agent Simulation of World Wars》(2023) GitHub: github.com/agiresearch/WarAgent [fig1]
《Cure the headache of Transformers via Collinear Constrained Attention》(2023) GitHub: github.com/codefuse-ai/Collinear-Constrained-Attention [fig3]
《SmoothLLM: Defending LLMs Against Jailbreaking Attacks》(2023) GitHub: github.com/arobey1/smooth-llm
《HiFiHR: Enhancing 3D Hand Reconstruction from a Single Image via High-Fidelity Texture》(2023) GitHub: github.com/viridityzhu/HiFiHR
《ParroT: Translating During Chat Using Large Language Models》(2023) GitHub: github.com/wxjiao/InstructMT

发布于 北京