PsyCoMark: Conspiracy Marker Extraction
Models mark the words that describe conspiracy claims in Reddit comments. The organizers score overlap with human annotations across five marker types.
Teams tested Qwen3 and Gemma 3. The organizers report 30 teams on CodaBench and publish scores for 16 qualifying teams.
- Made by
- PsyCoMark organizers
- Released
- March 2025
- Kind
- Reading comprehension
- Level
- School
- Scored
- Checked automatically
- Answers
- Short answer
- Tasks
- Public, plus a private test set
- Measures
- Capability
- Subject
- Media and communications
- Language
- English
- Interaction
- Single turn
- Items from
- Real work and use
- Refresh
- One fixed set
- Run by
- A host
HU tops the leaderboard with 0.26
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. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.
- 1 HU 0.26
- 2 CCNU 0.24
- 3 CredenceAI 0.24
- 4 dangphuduy 0.24
- 5 UMUTeam 0.24
- 6 UCSC NLP 0.22
- 7 AILS-NTUA 0.21
- 8 NUST PsyAI 0.21
- 9 YNU-HPCC 0.21
- 10 Team A 0.19
- 11 CUET_320 0.18
- 12 CuriosAI 0.18
- 13 wangkongqiang 0.15
- 14 zhangpeng 0.14
- 15 Jia 0.08
- 16 AGAI 0.07
Source: The PsyCoMark: Conspiracy Marker Extraction paper
| # | Entry | Score | Setting | Submitted by | Date | Source |
|---|---|---|---|---|---|---|
| 1 | HU | 0.26 | HU | Official leaderboard | ||
| 2 | CCNU | 0.24 | CCNU | Official leaderboard | ||
| 3 | CredenceAI | 0.24 | CredenceAI | Official leaderboard | ||
| 4 | dangphuduy | 0.24 | dangphuduy | Official leaderboard | ||
| 5 | UMUTeam | 0.24 | UMUTeam | Official leaderboard | ||
| 6 | UCSC NLP | 0.22 | UCSC NLP | Official leaderboard | ||
| 7 | AILS-NTUA | 0.21 | AILS-NTUA | Official leaderboard | ||
| 8 | NUST PsyAI | 0.21 | NUST PsyAI | Official leaderboard | ||
| 9 | YNU-HPCC | 0.21 | YNU-HPCC | Official leaderboard | ||
| 10 | Team A | 0.19 | Team A | Official leaderboard | ||
| 11 | CUET_320 | 0.18 | CUET_320 | Official leaderboard | ||
| 12 | CuriosAI | 0.18 | CuriosAI | Official leaderboard | ||
| 13 | wangkongqiang | 0.15 | wangkongqiang | Official leaderboard | ||
| 14 | zhangpeng | 0.14 | zhangpeng | Official leaderboard | ||
| 15 | Jia | 0.08 | Jia | Official leaderboard | ||
| 16 | AGAI | 0.07 | AGAI | Official 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 Knowledge
- MMLU-Pro In use Harder 10-option version of MMLU with reasoning-focused questions across 14 subjects.
Intern-S2-397B 89.8%
- MMLU In use 57-subject multiple-choice exam questions from elementary to professional level. Scored by accuracy.
Seed1.8 92.3%
- SuperGPQA In use 26,529 graduate-level multiple-choice questions across 285 disciplines. Accuracy.
Qwen 3.6 Max (Preview) 73.9%
- WideSearch In use 200 bilingual broad info-seeking tasks: agent collects many atomic facts into a table. Scored by success rate and row/item F1.
Tencent Hy4 preview 83.9%
- BBEH micro average In use Models solve harder replacements for each of the 23 BIG-Bench Hard tasks, 4,520 examples in total. The score is the share of all examples answered correctly.
Gemma 4 31B IT 74.4%
- HLE-Diamond In use Models answer 1,000 cleaned expert questions from Humanity's Last Exam, 500 on reasoning and 500 on knowledge, closed book. Graded on exact answers. Claude Fable 5.2 77.9%
Data audit
All 8 checks pass. Every benchmark
Catalog record 2 pass
- Pass Task access Public tasks, plus a private test set.
- Pass Confidence Checked against a primary source.
Scores 2 pass 3 notes
- Pass Independent runs Scores from the benchmark’s paper (16 entries).
- Pass Models scored 16 entries on its leaderboard, none a model we track.
- Note Newest score No scores yet.
- Note Frontier models No models scored yet.
- Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
- Pass Importer We read the board on codabench.org: 16 entries at the last read, none a model we track.
- 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.