tensor.news
Integrity IndexModelsCompareLabsPapersMethodologyAbout
tensor.news

© 2026

Evidence

  • Integrity Index
  • Model Claim Ledger
  • Compare

Entities

  • Models
  • Benchmarks
  • Labs
  • Papers

Method

  • Methodology
  • About
  • Privacy

Machine

  • Feed (Atom)
  • llms.txt
  • llms-full.txt
  • Sitemap

tensor.news grades benchmarks and tags model claims by evidence status. Every fact carries a source record and provenance. Scores are framed as task performance under a disclosed harness, never as deployed capability.

← Integrity Index

PostTrainBench

PostTrainBench

Rank #20 of 51 · 19 models scored · top score 34.29

AINTEGRITY 89 / 100

unknown

Strongest on contamination resistance, weakest on discrimination. Caveat: the public test set leaves contamination on the table.

consistent harnesspublic test set

Headroom to ceiling

65.7 headroomtop 34.3ceiling 100

Score distribution

median 17.8GLM-5.2 34.362136

State of the art over time

6213634.320252026

Integrity breakdown

Weighted composite. Each signal is computed from the source-backed leaderboard below.

Discrimination30%68

Does it still separate models?

Saturation headroom30%96

How far from ceiling / clustered at the top?

Contamination resistance15%100

Public vs held-out; training-leak risk.

Harness comparability15%100

Apples-to-apples, or mixed / vendor-optimized?

Freshness10%100

How old is the benchmark?

Leaderboard

#ModelScoreEvidenceMeasured
1GLM-5.234.29unverified2026-06-16
2Claude Opus 4.834.08unverified2026-05-28
3Claude Opus 4.728.56unverified2026-04-16
4GPT-5.525.02unverified2026-04-23
5Claude Opus 4.624.82unverified2026-02-05
6Gemini 3.1 Pro21.59unverified2026-02-19
7GPT-5.221.38unverified2025-12-11
8GPT-5.420.23unverified2026-03-05
9Gemini 3 Pro18.12unverified2025-11-18
10GPT-5.3 Codex17.76unverified2026-02-05
11Claude Opus 4.517.29unverified2025-11-24
12Claude Sonnet 4.616.42unverified2026-02-17
13GLM-513.88unverified2026-02-11
14Kimi K2.510.26unverified2026-02-02
15Claude Sonnet 4.59.94unverified2025-09-29
16MiniMax-M2.59.5unverified2026-02-12
17GLM-4.77.48unverified2025-12-22
18Qwen3-Max7.42unverified2025-09-05
19Kimi K2 Thinking7.25unverified2025-11-06

At a glance

Models scored
19
Top model
GLM-5.2
Top score
34.29
Score spread
27.04
Headroom
65.71

Benchmark context

Task type
unknown
Metric
normalized performance in Epoch CSV, rendered as accuracy (%)
Score ceiling
100
Human baseline
unknown
Test-set privacy
unknown
Construction
unknown
Key assumptions
unknown
Limitations
unknown
Contamination history
unknown
Harness notes
Epoch CSV reports model, model version, performance, optimized flag, score source, benchmark release date, and model date.
Authors
unknown
Released
unknown
License
CC BY 4.0
Source
epoch.ai/data/eci_benchmarks.csv

Provenance

19 source records · scores from epoch_benchmarks. Every fact on this page is evidence-linked.

Follow the record

Integrity neighbors

  • BalrogA · 89
  • ExploitBenchA · 89
  • GeoBenchA · 89
  • OTIS Mock AIME 2024-2025A · 88
  • WeirdMLA · 88
  • Aider polyglotA · 87

Top models here

  • GLM-5.234.29
  • Claude Opus 4.834.08
  • Claude Opus 4.728.56
  • GPT-5.525.02
  • Claude Opus 4.624.82

Head-to-head

  • GLM-5.2 vs Claude Opus 4.8
  • GLM-5.2 vs Claude Opus 4.7
  • Claude Opus 4.8 vs Claude Opus 4.7