bytedance-research/pasa-7b-selector

TEXT GENERATIONConcurrency Cost:1Model Size:7.6BQuant:FP8Ctx Length:32kLicense:cc-by-nc-sa-4.0Architecture:Transformer0.0K Open Weights Cold

The bytedance-research/pasa-7b-selector is a 7.6 billion parameter language model developed by Bytedance Research, designed as an LLM agent for comprehensive academic paper search. This model specializes in navigating and understanding academic literature, offering a unique capability for researchers. With a substantial context length of 131072 tokens, it is optimized for processing extensive textual information relevant to scientific and scholarly documents.

Loading preview...

PaSa: An LLM Agent for Comprehensive Academic Paper Search

PaSa (Paper Search Agent) is a 7.6 billion parameter language model developed by Bytedance Research, specifically engineered to function as an intelligent agent for academic paper search. Unlike general-purpose LLMs, PaSa is tailored to understand, process, and retrieve information from scholarly articles, making it a specialized tool for academic research.

Key Capabilities

  • Academic Paper Search: Designed to assist users in finding relevant academic papers based on queries.
  • Large Context Window: Features a significant context length of 131072 tokens, enabling it to handle and analyze extensive academic texts.
  • LLM Agent Architecture: Operates as an agent, suggesting its ability to perform multi-step reasoning and interaction within the search process.

Good For

  • Researchers: Ideal for academics, students, and professionals who frequently need to navigate and extract information from scientific literature.
  • Literature Review: Can significantly streamline the process of conducting comprehensive literature reviews.
  • Information Retrieval: Specialized in retrieving precise information from academic databases and papers.

For more technical details, refer to the associated paper: PaSa: An LLM Agent for Comprehensive Academic Paper Search.