https://github.com/fangyuan-ksgk/CoT-Reasoning-without-Prompting\nInsight: Decoding time reasoning is cheap, effective, and can bring out the 'inherent' reasoning capacity from pre-trained LLM. \nDrawback: Indentification of the set of answer, and its location reamains the million dollar question. ","text":"Our Re-Implementation code: https://github.com/fangyuan-ksgk/CoT-Reasoning-without-Prompting\nInsight: Decoding time reasoning is cheap, effective, and can bring out the 'inherent' reasoning capacity from pre-trained LLM. \nDrawback: Indentification of the set of answer, and its location reamains the million dollar question. "},"id":"2402.10200","title":"Chain-of-Thought Reasoning Without Prompting","thumbnailUrl":"https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2402.10200.png","upvotes":109,"publishedAt":"2024-02-15T18:55:41.000Z","isUpvotedByUser":false},{"_id":"65dd5c2fec8083e4ab70815f","position":2,"gallery":["https://cdn-uploads.huggingface.co/production/uploads/64d98ef7a4839890b25eb78b/iBeBaAXbhpDauIXP3PnNY.png"],"type":"paper","note":{"html":"In-Context-Learning based preference alignment, performance on-par with Supervised Fine-Tuning (SFT). Can be used to generated optimal preference pairs, or augment the preference dataset.","text":"In-Context-Learning based preference alignment, performance on-par with Supervised Fine-Tuning (SFT). Can be used to generated optimal preference pairs, or augment the preference dataset."},"id":"2402.09320","title":"ICDPO: Effectively Borrowing Alignment Capability of Others via\n In-context Direct Preference Optimization","thumbnailUrl":"https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2402.09320.png","upvotes":6,"publishedAt":"2024-02-14T17:14:34.000Z","isUpvotedByUser":false},{"_id":"65dd5c4c29b761fff4be0aa2","position":3,"gallery":["https://cdn-uploads.huggingface.co/production/uploads/64d98ef7a4839890b25eb78b/ZgFz-zf_kYmEhhDo5X_io.png"],"type":"paper","note":{"html":"Self-Discover solves any task in three steps: \nPickging a reasoning structure, designing a stepwise reasoning plan, then implement the thinking process to get the answer. \nSignificant performance improvement is observed. ","text":"Self-Discover solves any task in three steps: \nPickging a reasoning structure, designing a stepwise reasoning plan, then implement the thinking process to get the answer. \nSignificant performance improvement is observed. "},"id":"2402.03620","title":"Self-Discover: Large Language Models Self-Compose Reasoning Structures","thumbnailUrl":"https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2402.03620.png","upvotes":117,"publishedAt":"2024-02-06T01:13:53.000Z","isUpvotedByUser":false}],"position":0,"theme":"orange","private":false,"shareUrl":"https://hf.co/collections/Temus/foundation-ai-papers","upvotes":35,"isUpvotedByUser":false},{"slug":"Temus/foundation-models-and-tools-65de066dc78f70e147b8367e","title":"Foundation Models and Tools","description":"","gating":false,"lastUpdated":"2024-07-18T00:14:13.795Z","owner":{"_id":"65dd51900481e2a395cb0550","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/64d98ef7a4839890b25eb78b/zY8Q9FVxDOkcwAyY-ckXg.png","fullname":"Temus","name":"Temus","type":"org","isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":16,"isUserFollowing":false},"items":[{"_id":"65de067670202c2ac1bd5f0d","position":0,"type":"paper","id":"2402.10986","title":"FinTral: A Family of GPT-4 Level Multimodal Financial Large Language\n Models","thumbnailUrl":"https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2402.10986.png","upvotes":81,"publishedAt":"2024-02-16T05:05:12.000Z","isUpvotedByUser":false},{"_id":"65e06c480123d090c9ad547e","position":1,"type":"model","author":"bigcode","authorData":{"_id":"62ce8f4248fbe688600093a0","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1659521200179-5e48005437cb5b49818287a5.png","fullname":"BigCode","name":"bigcode","type":"org","isHf":false,"isHfAdmin":false,"isMod":false,"plan":"team","followerCount":1886,"isUserFollowing":false},"downloads":39230,"gated":false,"id":"bigcode/starcoder2-15b","availableInferenceProviders":[],"lastModified":"2024-06-05T19:52:45.000Z","likes":662,"pipeline_tag":"text-generation","private":false,"repoType":"model","isLikedByUser":false,"widgetOutputUrls":[]},{"_id":"65e06cc31d8870833f76e6e4","position":2,"type":"paper","note":{"html":"Zephyr is by far the best aligned open-sourced LLM I've used. They recently have a -beta and a -gamma (fine-tuned out of Gemma) version too.","text":"Zephyr is by far the best aligned open-sourced LLM I've used. They recently have a -beta and a -gamma (fine-tuned out of Gemma) version too."},"id":"2310.16944","title":"Zephyr: Direct Distillation of LM Alignment","thumbnailUrl":"https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2310.16944.png","upvotes":123,"publishedAt":"2023-10-25T19:25:16.000Z","isUpvotedByUser":false},{"_id":"65e1f873c368fce1337332cc","position":3,"type":"model","author":"mixedbread-ai","authorData":{"_id":"6579b8e83db8c022afd86d63","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/643ee0870d1194da249bd7fe/voYdYlFgQH5vyMwyMNluZ.png","fullname":"Mixedbread","name":"mixedbread-ai","type":"org","isHf":false,"isHfAdmin":false,"isMod":false,"plan":"team","followerCount":344,"isUserFollowing":false},"downloads":22749,"gated":false,"id":"mixedbread-ai/mxbai-rerank-large-v1","availableInferenceProviders":[],"lastModified":"2025-04-02T14:41:49.000Z","likes":139,"pipeline_tag":"text-ranking","private":false,"repoType":"model","isLikedByUser":false}],"position":1,"theme":"green","private":false,"shareUrl":"https://hf.co/collections/Temus/foundation-models-and-tools","upvotes":1,"isUpvotedByUser":false},{"slug":"Temus/llm-evaluation-65e525d95132c2edd117acce","title":"LLM-evaluation","description":"Evaluation of LLM agents paper","gating":false,"lastUpdated":"2024-09-05T07:57:22.673Z","owner":{"_id":"65dd51900481e2a395cb0550","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/64d98ef7a4839890b25eb78b/zY8Q9FVxDOkcwAyY-ckXg.png","fullname":"Temus","name":"Temus","type":"org","isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":16,"isUserFollowing":false},"items":[{"_id":"65f32e318ac970331d38482e","position":0,"type":"paper","id":"2401.15391","title":"MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop\n Queries","thumbnailUrl":"https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2401.15391.png","upvotes":6,"publishedAt":"2024-01-27T11:41:48.000Z","isUpvotedByUser":false},{"_id":"6605960a196d49210e5e719c","position":1,"type":"paper","id":"2403.18802","title":"Long-form factuality in large language models","thumbnailUrl":"https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2403.18802.png","upvotes":26,"publishedAt":"2024-03-27T17:48:55.000Z","isUpvotedByUser":false},{"_id":"660926c8df88d3dbacd3045e","position":2,"type":"paper","note":{"html":"Evaluator local LLM fine-tuned to mimic GPT-4 performance\n1. Swap augmentation\n2. Scenario additional augmentation (Rational Prompt)","text":"Evaluator local LLM fine-tuned to mimic GPT-4 performance\n1. Swap augmentation\n2. Scenario additional augmentation (Rational Prompt)"},"id":"2310.17631","title":"JudgeLM: Fine-tuned Large Language Models are Scalable Judges","thumbnailUrl":"https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2310.17631.png","upvotes":35,"publishedAt":"2023-10-26T17:48:58.000Z","isUpvotedByUser":false},{"_id":"660926d88db1f6ae03d7fccb","position":3,"type":"paper","id":"2310.08491","title":"Prometheus: Inducing Fine-grained Evaluation Capability in Language\n Models","thumbnailUrl":"https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2310.08491.png","upvotes":57,"publishedAt":"2023-10-12T16:50:08.000Z","isUpvotedByUser":false}],"position":2,"theme":"indigo","private":false,"shareUrl":"https://hf.co/collections/Temus/llm-evaluation","upvotes":1,"isUpvotedByUser":false},{"slug":"Temus/foundation-ai-papers-ii-663041c21ca040c84d992cc8","title":"Foundation AI Papers (II)","description":"","gating":false,"lastUpdated":"2024-07-23T10:45:51.495Z","owner":{"_id":"65dd51900481e2a395cb0550","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/64d98ef7a4839890b25eb78b/zY8Q9FVxDOkcwAyY-ckXg.png","fullname":"Temus","name":"Temus","type":"org","isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":16,"isUserFollowing":false},"items":[{"_id":"663263d6a2354b0f50e17036","position":0,"type":"paper","id":"2404.19733","title":"Iterative Reasoning Preference Optimization","thumbnailUrl":"https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2404.19733.png","upvotes":49,"publishedAt":"2024-04-30T17:28:05.000Z","isUpvotedByUser":false},{"_id":"663263df48f6e5036f5909dd","position":1,"type":"paper","note":{"html":"well ...","text":"well ..."},"id":"2404.19737","title":"Better & Faster Large Language Models via Multi-token Prediction","thumbnailUrl":"https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2404.19737.png","upvotes":81,"publishedAt":"2024-04-30T17:33:57.000Z","isUpvotedByUser":false},{"_id":"663263f7d5677e32879e5ba4","position":2,"type":"paper","id":"2403.07691","title":"ORPO: Monolithic Preference Optimization without Reference Model","thumbnailUrl":"https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2403.07691.png","upvotes":72,"publishedAt":"2024-03-12T14:34:08.000Z","isUpvotedByUser":false},{"_id":"6632ddc12ba96ac6ee91033d","position":3,"type":"paper","note":{"html":"\"Less scalable version\" of AGI backend model","text":"\"Less scalable version\" of AGI backend model"},"id":"2404.19756","title":"KAN: Kolmogorov-Arnold Networks","thumbnailUrl":"https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2404.19756.png","upvotes":116,"publishedAt":"2024-04-30T17:58:29.000Z","isUpvotedByUser":false}],"position":3,"theme":"pink","private":false,"shareUrl":"https://hf.co/collections/Temus/foundation-ai-papers-ii","upvotes":4,"isUpvotedByUser":false},{"slug":"Temus/planning-with-llm-667d06cc7bd463a3909377b5","title":"Planning-with-LLM","gating":false,"lastUpdated":"2024-09-04T11:27:29.960Z","owner":{"_id":"65dd51900481e2a395cb0550","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/64d98ef7a4839890b25eb78b/zY8Q9FVxDOkcwAyY-ckXg.png","fullname":"Temus","name":"Temus","type":"org","isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":16,"isUserFollowing":false},"items":[{"_id":"667d06cc7bd463a3909377b6","position":0,"type":"paper","id":"2406.18082","title":"Octo-planner: On-device Language Model for Planner-Action Agents","thumbnailUrl":"https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2406.18082.png","upvotes":48,"publishedAt":"2024-06-26T05:40:10.000Z","isUpvotedByUser":false},{"_id":"668242e5224478a1f6a7002f","position":1,"type":"paper","id":"2406.13892","title":"Adaptable Logical Control for Large Language Models","thumbnailUrl":"https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2406.13892.png","upvotes":1,"publishedAt":"2024-06-19T23:47:59.000Z","isUpvotedByUser":false},{"_id":"668243e595989c5c95679373","position":2,"type":"paper","id":"2406.19215","title":"SeaKR: Self-aware Knowledge Retrieval for Adaptive Retrieval Augmented\n Generation","thumbnailUrl":"https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2406.19215.png","upvotes":32,"publishedAt":"2024-06-27T14:38:33.000Z","isUpvotedByUser":false},{"_id":"6682b4520b54a28aff06cbd8","position":3,"type":"paper","note":{"html":"RAG do two things. Slicing & Retrieving. Slicing document into passages ignores the connection between passage, and the resulting \"database\" is not structured. Human stored their knowledge with more structure, related information of one concept are usually clustered together. HippoRAG adopts a LLM to process the knowledge base into a graph structure, continuously updating the knowledge graph with new information flow. During retrieving, most relevant nodes along with their neighbours are picked ","text":"RAG do two things. Slicing & Retrieving. Slicing document into passages ignores the connection between passage, and the resulting \"database\" is not structured. Human stored their knowledge with more structure, related information of one concept are usually clustered together. HippoRAG adopts a LLM to process the knowledge base into a graph structure, continuously updating the knowledge graph with new information flow. During retrieving, most relevant nodes along with their neighbours are picked "},"id":"2405.14831","title":"HippoRAG: Neurobiologically Inspired Long-Term Memory for Large Language\n Models","thumbnailUrl":"https://cdn-thumbnails.huggingface.co/social-thumbnails/papers/2405.14831.png","upvotes":5,"publishedAt":"2024-05-23T17:47:55.000Z","isUpvotedByUser":false}],"position":4,"theme":"orange","private":false,"shareUrl":"https://hf.co/collections/Temus/planning-with-llm","upvotes":2,"isUpvotedByUser":false}],"datasets":[],"models":[],"paperPreviews":[],"spaces":[],"buckets":[],"numBuckets":0,"numDatasets":0,"numModels":0,"numSpaces":1,"lastOrgActivities":[],"acceptLanguages":["*"],"canReadRepos":false,"canReadSpaces":false,"blogPosts":[],"currentRepoPage":0,"filters":{},"paperView":false}">
Edit this README.md markdown file to author your organization card.