AI Paper Atlas

Research map / Large language models / Agents, code and math

DEER: A Benchmark for Evaluating Deep Research Agents on Expert Report Generation

Janghoon Han, Heegyu Kim, Changho Lee, Dahm Lee, Min Hyung Park, Hosung Song, Stanley Jungkyu Choi, Moontae Lee, Honglak Lee

ICML 2026 · Large language models › Agents, code and math · approach: LLMs for scientific discovery · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add reasoning (think) to general LLMsdeep research benchmark systems · DEER overall meanmean score: 4.88 -> 5.89 (GPT-5.2 fast -> think)Improved
add web search to reasoning LLMsGPT-5.2 · DEER overall meanmean score: 5.89 -> 6.62Improved
add web search or deep research orchestration to reasoning modelsreport-quality dimensions excluding information scores · DEERreport-quality scoresHurt
add dimensions to the vanilla LLM judge (ablation)LLM-as-judge · human correlationPearson r: 0.63 -> 0.66No clear effect
add 101 granular rubric items (ablation)LLM-as-judge with dimensions · human correlationPearson r: 0.66 -> 0.64Hurt
add task-specific expert guidance (ablation)LLM-as-judge with granular rubrics · human correlationPearson r: 0.64 -> 0.73Improved
add expert guidance (ablation)LLM-as-judge · inter-evaluator reliabilityKrippendorff alpha: 0.33 -> 0.55Improved
add dimensions to vanilla judge (ablation)LLM-as-judge · inter-evaluator reliabilityKrippendorff alpha: 0.46 -> 0.32Hurt
use 10-point scale instead of 5-point scale (ablation)report quality judge · human correlationPearson r: 0.694 -> 0.734Improved
add 5-band score anchors (ablation)report quality judge · human correlationPearson r: 0.722 -> 0.734Improved
semantic back-tracking vs sliding window for citation recovery (ablation)information verification · B/C claimsJaccard: 0.6822 -> 0.7070Improved
grouped verification by evidence URL (ablation)claim verification (GPT-5-mini) · claim verificationF1: 87.25 -> 90.91; cost 12.84 -> 3.65Improved

Later papers that compared with this method

15 later accepted papers report DEER as a baseline in their results tables; 14 of them report a better number on most of the results they share with it.

7 more not listed.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy