Question-type-specific LLM pipeline boosts BioASQ 14b biomedical QA
A new framework for BioASQ 14b Task B selects different inference procedures for yes/no, factoid, and list questions, improving answer robustness and evidence grounding. The approach uses question-type-specific prompting strategies rather than a single method for all queries.
Entities
Related
RSF-GLLM framework decouples differentiable graph reasoning from answer generation for multi-hop QA over knowledge graphs
Researchers propose RSF-GLLM, a framework that decouples differentiable graph reasoning from answer generation to address the non-differentiability issue in traditional retrieve-then-read pipelines for multi-hop question answering over knowledge graphs. The Recurrent Soft-Flow module uses a GRU-guided query updater and dynamic gating to propagate relevance scores across semantically dissimilar bridge nodes.
DynaKRAG: A unified framework for learnable evidence control in multi-hop RAG
Researchers propose DynaKRAG, a framework that learns a state-conditioned policy to dynamically choose among evidence operations (retrieval, reformulation, critique, sufficiency checking) in multi-hop retrieval-augmented generation. This moves beyond fixed pipelines, potentially improving accuracy and flexibility in complex QA tasks.
DataGovBench: New benchmark evaluates LLMs on real-world data analysis with large multi-tabular datasets
Researchers introduced DataGovBench, a benchmark derived from governmental open data to evaluate LLMs on practical data analysis tasks. It includes Table QA and Table Insight tasks, addressing limitations of existing benchmarks that focus on small tables and fact retrieval.
Billy Bassistant AI Fish: AI assistant with local file knowledge and status LED
Billy Bassistant AI Fish is a voice-enabled AI assistant that can be fed custom PDF, Excel, or text files organized into topic-specific folders. It uses LLMs to process these local files and answer questions based on them, and optionally shows a status LED indicator. It solves the problem of having a physical AI assistant that can access and reason over personal documents.
LLM hardware recipe database with filters and community usage tracking
A community-driven database that lists which LLM models run on which hardware, with performance details. Users can filter by hardware, submit new recipes, and mark which recipes they actively use to show popularity. It solves the problem of finding compatible model-hardware combinations for LLM deployment.