DiscourseEE
Models read Reddit discussions about opioid use disorder and extract details about medication use, tapering, and relapse. Some details require inference across a conversation.
Two later papers with overlapping authors used DiscourseEE. REGen evaluated 98 test documents. A data augmentation study evaluated only examples with explicit text spans.
- Made by
- Dartmouth College
- Released
- October 2024
- Kind
- Medicine
- Level
- University
- Scored
- Checked automatically
- Answers
- Short answer
- Tasks
- All public
- Measures
- Capability
- Subject
- Clinical medicine
- Language
- English
- Interaction
- Single turn
- Items from
- Real work and use
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Biology and medicine
- Strongest model tested
-
Llama 3-70B #182 on the capability index
- Last model added
- April 2024
- Scores refreshed
- November 20th, 2026
GPT-4 (snapshot unspecified) tops the leaderboard with 42% Llama 3-8B leads with 36.9% among models released by April 2024
Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is Llama 3-8B, at 36.9%. This view shows only models we track, each at its best.
The newest model with a score here came out in April 2024, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 1 GPT-4 (snapshot unspecified) question-guided prompt 42%
- 2
Llama 3-8B description-guided prompt 36.9%
- 3
Llama 3-70B question-guided prompt 36.3%
- 4 FLAN-T5 (Large) description-guided prompt 35.5%
- 5
Mixtral 8x7B description-guided prompt 35.5%
- 6
Gemma 7B question-guided prompt 30.2%
- 7 FLAN-T5 (Base) description-guided prompt 26.4%
- 8 Extractive-QA description-guided prompt 17.1%
Source: The DiscourseEE paper
- 1
Llama 3-8B description-guided prompt 36.9%
- 2
Llama 3-70B question-guided prompt 36.3%
- 3
Mixtral 8x7B description-guided prompt 35.5%
- 4
Gemma 7B question-guided prompt 30.2%
Source: The DiscourseEE paper
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-4 (snapshot unspecified) | 42% | question-guided prompt | Official leaderboard | |
| 2 | | 36.9% | description-guided prompt | Apr 2024 | Official leaderboard |
| 3 | | 36.3% | question-guided prompt | Apr 2024 | Official leaderboard |
| 4 | FLAN-T5 (Large) | 35.5% | description-guided prompt | Official leaderboard | |
| 5 | | 35.5% | description-guided prompt | Dec 2023 | Official leaderboard |
| 6 | | 30.2% | question-guided prompt | Feb 2024 | Official leaderboard |
| 7 | FLAN-T5 (Base) | 26.4% | description-guided prompt | Official leaderboard | |
| 8 | Extractive-QA | 17.1% | description-guided prompt | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 36.9% | description-guided prompt | Apr 2024 | Official leaderboard |
| 2 | | 36.3% | question-guided prompt | Apr 2024 | Official leaderboard |
| 3 | | 35.5% | description-guided prompt | Dec 2023 | Official leaderboard |
| 4 | | 30.2% | question-guided prompt | Feb 2024 | Official leaderboard |
Scores at each effort level
Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.
Source: The DiscourseEE paper
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| description-guided prompt | 36.9% | ||||
| question-guided prompt | 36.3% | ||||
| description-guided prompt | 35.8% | ||||
| description-guided prompt | 35.5% | ||||
| question-guided prompt | 35% | ||||
| question-guided prompt | 32.4% | ||||
| question-guided prompt | 30.2% | ||||
| description-guided prompt | 29.9% |
Scores at each effort level
Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.
Source: The DiscourseEE paper
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| description-guided prompt | 36.9% | ||||
| question-guided prompt | 36.3% | ||||
| description-guided prompt | 35.8% | ||||
| description-guided prompt | 35.5% | ||||
| question-guided prompt | 35% | ||||
| question-guided prompt | 32.4% | ||||
| question-guided prompt | 30.2% | ||||
| description-guided prompt | 29.9% |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: The DiscourseEE paper
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official DiscourseEE leaderboard
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Law, finance and medicine
- MedXpertQA-MM In use Models answer expert-level medical exam questions that combine images with clinical records and test results. Scoring is accuracy.
Qwen 3.8 Max 80.4%
- HealthBench Professional In use 525 physician-authored tasks from real clinician chats (care consult, documentation, research) with physician-adjudicated rubrics. Claude Sonnet 5.5 69.2%
- HealthBench In use 5,000 realistic health conversations graded against physician-written rubrics. A length-adjusted variant penalizes verbosity. GPT-5 67.2%
- FinSearchComp In use Agents search the web to fetch live financial data, look up history, and run multi-step investigations, across 635 expert-written questions. Scored by answer accuracy.
Ling-3.1-flash 81.7%
- KorMedMCQA In use Models answer 7,469 multiple-choice questions from Korean licensing exams for doctors, nurses, pharmacists and dentists, held from 2012 to 2024. o1-preview 92.7%
- Harvey LAB-AA v1.1 In use Artificial Analysis's run of Harvey's Legal Agent Benchmark on Harvey's newer private set of 120 tasks across 24 practice areas. A deliverable with a material hallucination scores zero, and the headline is the share of judges who find every rubric criterion passed. GPT-6.2 Astra 15%
Data audit
9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 2 years ago, in April 2024, for Llama 3-70B. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
- Pass Independent runs Scores from the benchmark’s paper (8 entries).
- Pass Models scored 4 models we track, plus 4 other entries on its leaderboard.
- Warning Newest score The newest came 2 years ago, in April 2024, for Llama 3-70B.
- Note Frontier models None of the top 10 has a score. The strongest tested is Llama 3-70B (#182).
- Pass Headroom The best is 36.9%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on aclanthology.org: 8 rows for models we track and 4 other entries at the last read.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.