SACHIN KUMARCreating Dynamic Mindmap of information from collaborative discourse among LLM Agents with…LLM powered chatbots and generative search engines excel at answering concrete queries, discovering information in the terrain of unknown…4d ago4d ago
SACHIN KUMARMegaAgent: framework for autonomous cooperation in Large-Scale LLM Agent SystemsLLM powered multiagent(LLM-MA) systems mostly follow predefined Standard Operating Procedures(SOPs) that remain unchanged across the whole…Aug 22Aug 22
SACHIN KUMARRAGLAB: Modular and Research-Oriented Unified Framework for Retrieval-Augmented GenerationIn this paper[1], authors introduce RAGLAB, a modular and research-oriented open-source library. RAGLAB reproduces 6 existing algorithms…Aug 22Aug 22
SACHIN KUMARImportance Weighting-based Self-Improvement : framework to enhance effectiveness of LLM…Typical paradigm of LLM self-improvement involves training LLM on self-generated data, part of which may be detrimental and should be…Aug 21Aug 21
SACHIN KUMARMeta Agent Search : Automated System Design Creation for LLM based Agentic SystemsIn this paper[1], authors formulate a new research area, Automated Design of Agentic Systems (ADAS), which aims to automatically create…Aug 19Aug 19
SACHIN KUMAREfficientRAG: an efficient retriever for multi-hop question answeringIn this paper[1], authors introduce EfficientRAG, an efficient retriever for multi-hop question answering. EfficientRAG iteratively…Aug 18Aug 18
SACHIN KUMARI-SHEEP: Self-Alignment of LLMs from Scratch via an Iterative Self-Enhancement ParadigmIn this paper[1], authors introduce I-SHEEP, an Iterative Self-EnHancEmEnt Paradigm. This human-like paradigm enables LLMs to continuously…Aug 17Aug 17
SACHIN KUMARToolSandbox: Stateful, Conversational, Interactive Evaluation Benchmark for LLM Agents Tool Use…In this paper[1], authors present TOOLSANDBOX, which includes stateful tool execution, implicit state dependencies between tools, a…Aug 15Aug 15
SACHIN KUMARLongWriter : using LLM Agent-based pipeline to scale LLMs output window size to 10,000+ wordsLLMs can process inputs up to 100,000 tokens, yet struggle to generate outputs exceeding even a modest length of 2,000 words as model’s…Aug 14Aug 14
SACHIN KUMARA Hybrid RAG System with Optimizations for Complex ReasoningIn this paper [1], authors introduce a novel Retrieval-Augmented Generation (RAG) system designed for real-world applications. Proposed…Aug 13Aug 13