ClaimWriter cuts deep research hallucination up to 4.5×
Faithfulness can be fixed at the drafting stage rather than by better retrieval — the writer, not the search, is where most hallucinations enter. Teams building deep research pipelines can plug ClaimWriter into existing frameworks and audit outputs with ClaimProbe using the released code. The benchmark offers a reproducible baseline for measuring claim-level grounding across competing agent stacks.
ClaimWriter restructures how deep research agents write reports: it extracts facts from sources, maps them to an outline derived from the user's query, and drafts each section from source-linked claims. Swapping only the writer into three frameworks — Enterprise Deep Research, NVIDIA AI-Q and Open Deep Research — with search fixed produced 2.6–4.5× less hallucination and 1.2–1.7× better recall of necessary facts. Overall report quality was largely preserved. The team also released ClaimProbe, a claim-level audit that checks every report claim against retrieved evidence, with code and benchmark publicly available.
Redesigning and auditing deep research writing for faithful reports: https://t.co/kZXAiqs78z
Deep research agents can write fluent, well-scored reports that still miss key facts, cite the wrong sources, or make claims no source supports.
ClaimWriter changes how the report gets written. It extracts facts from sources, maps them to an outline built from the user's query, and drafts each section from source-linked claims.
We swapped only the report writer into three deep research frameworks (Enterprise Deep Research, NVIDIA AI-Q, and Open Deep Research), keeping search and evidence collection fixed: → 2.6–4.5× less hallucination → 1.2–1.7× better recall of necessary facts → Overall report quality largely preserved
To measure the difference, we built ClaimProbe, a claim-level audit that checks each report claim against the retrieved evidence.
Code and benchmark: https://t.co/kF5EFZ8FQq
#FutureOfAI #EnterpriseAI
