Scalability of CIBER's Evidence Retrieval and Its Impact on Scientific Claim Verification Accuracy with Up to 1000 Documents

Abstract

Abstract: Large Language Models(LLMs) now handle tasks like question answering, summarisation, code generation and dialogue with impressive results.Yet they still suffer from a key issue: hallucination happens when a model generates text that reads well but is factually wrong or not grounded in real evidence.The risk is higher in domains like healthcare, law, finance and research, where inaccurate outputs can lead to real damage.This survey focuses on how to detect hallucination verification in LLMs.We review 5 core detection approaches: retrieval-based, uncertainty-based, embedding-based, learning-base

Research Question

How does the scalability of CIBER's evidence retrieval mechanism affect the accuracy of scientific claim verification when integrating up to 1000 corroborating/refuting documents per claim?

Verification Level

Descriptive public verification status only; aggregate claim counts are public, but individual claim records are not exposed here.

Truth-Engine Gate Verdict

This record has not completed Gate 2 of the verification pipeline (a type-checked Lean4 proof for mathematical claims, or a sealed-sandbox reproduction for empirical claims). It is a literature synthesis only. VERIFIED requires an attached reproducible artifact (Lean4 proof source, or repro script and results) before this status can be set; it is not derived from review score or claim count.

Quality Tier

Descriptive public triage only; this tier does not alter current publication or DOI behavior.

Quality Dimensions

Automated triage signals derived from public fields; not human peer review or independent validation.

Correction Record

Public corrections are additive records. Current status does not claim the synthesis is error-free.

Provenance