Ai2 ScholarQA follows a RAG-based, multi-step prompting workflow using Claude Sonnet 3.5 model. It relies on a corpus of open-access papers. Ai2 ScholarQA is meant to satisfy literature searches that require insights from multiple relevant documents, and synthesize those insights into a comprehensive report. After receiving a query, the system first queries the index for the top k passages. These passages are further re-ranked with a pretrained transformer model and the top 50 candidates are retained for further processing. The answer generation is a 3-step process driven by prompts to an LLM: (1) quote extraction, (2) answer outline and clustering, (3) report generation. Find more details about the tool. [information updated on 2025.03.07]

Consensus uses language models (LLMs) and purpose-built search technology (Vector search) to surface relevant papers. The current source material used comes from the Semantic Scholar. Consensus run a custom, fine-tuned language model over the entire corpus of research papers and extract the Key Takeaway from every paper. Next, elements such as ‘what’, ‘is’ and ‘are’ are removed from the query and a combination of keyword search and Vector search is performed on the abstract and title of all articles. Find more details about the tool. [information updated on 2025.03.07]

Elicit searches across over 126 million academic papers from the Semantic Scholar corpus across all academic disciplines, including preprints (from arXiv, bioRxiv, and other repositories). It also includes non-open-access papers. Elicit answer will be based on the full text of the paper if it is available via open access, otherwise Elicit will answer based on just the abstract. Find more details about the tool. [information updated on 2025.03.07]

ORKG Ask employs a sophisticated pipeline to ensure you receive the most relevant information from the vast CORE dataset of research articles. The system uses an embedding model to encode your query and searches for semantically similar research articles using a vector store with an Approximate Nearest Neighbor (ANN) algorithm. ORKG Ask leverages the following advanced technologies to deliver its robust search and extraction capabilities: Qdrant vector store, Nomic embedding model and Mistral Instruct 7B v0.2 large language model. Find more details about the tool. [information updated on 2025.03.07]