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

CritPt

CritPt

Rank #14 of 51 · 47 models scored · top score 30.57

AINTEGRITY 92 / 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

69.4 headroomtop 30.6ceiling 100

Score distribution

median 2.9GPT-5.5 Pro 30.601632

State of the art over time

0163230.620242026

Integrity breakdown

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

Discrimination30%76

Does it still separate models?

Saturation headroom30%98

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
1GPT-5.5 Pro30.57unverified2026-04-23
2GPT-5.4 Pro30unverified2026-03-05
3Claude Fable 528.57unverified2026-06-09
4GPT-5.527.14unverified2026-04-23
5GPT-5.423.43unverified2026-03-05
6Claude Opus 4.820.86unverified2026-05-28
7GLM-5.220.86unverified2026-06-16
8Gemini 3.1 Pro17.71unverified2026-02-19
9Qwen3.7-Max13.43unverified2026-05-19
10Gemini 3.5 Flash13.14unverified2026-05-19
11DeepSeek-V4-Pro12.86unverified2026-04-24
12GPT-512.6unverified2025-08-07
13Claude Opus 4.712unverified2026-04-16
14Muse Spark11.33unverified2026-04-08
15Kimi K2.7 Code10unverified2026-06-12
16GPT-5.4 Mini10unverified2026-03-17
17GPT-5.4 Nano9.25unverified2026-03-17
18Grok 4.3 Beta8unverified2026-04-17
19Kimi K2.68unverified2026-04-20
20Gemini 3 Pro6.9unverified2025-11-18
21GPT-5.14.86unverified2025-11-13
22GLM-5.14.57unverified2026-04-07
23Claude Sonnet 4.63.14unverified2026-02-17
24Qwen 3.6 Plus2.86unverified2026-04-01
25Gemini 2.5 Pro (Jun 2025)2unverified2025-06-05
26o31.4unverified2024-12-20
27Gemini 3.1 Flash-Lite1.14unverified2026-03-03
28gpt-oss-120b1.14unverified2025-08-05
29Claude Sonnet 4.51.14unverified2025-09-29
30DeepSeek-R11.1unverified2025-01-20
31Gemini 2.5 Flash (Jun 2025)1.1unverified2025-06-17
32o4-mini0.6unverified2025-04-16
33MiniMax-M2.70.57unverified2026-03-18
34Claude Opus 40.3unverified2025-05-22
35Claude Sonnet 40.29unverified2025-05-22
36Gemma 3 27B0unverified2025-03-12
37Claude 3.5 Haiku0unverified2024-10-22
38Llama 3.1-8B0unverified2024-07-23
39Llama 4 Maverick0unverified2025-04-05
40Llama 4 Scout0unverified2025-04-05
41Mistral Small 3.10unverified2025-03-17
42Claude Haiku 4.50unverified2025-10-15
43DeepSeek-V30unverified2024-12-24
44DeepSeek-V3 (Mar 2025)0unverified2025-03-24
45GPT-4.1 mini0unverified2025-04-14
46GPT-4.1 nano0unverified2025-04-14
47GPT-4o (Nov 2024)0unverified2024-05-13

At a glance

Models scored
47
Top model
GPT-5.5 Pro
Top score
30.57
Score spread
30.57
Headroom
69.43

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

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

Follow the record

Integrity neighbors

  • GBAEvalA · 98
  • GDPvalA · 98
  • FrontierMath-Tier-4-2025-07-01-PrivateA · 93
  • The Agent CompanyA · 92
  • GSO-BenchA · 90
  • BalrogA · 89

Top models here

  • GPT-5.5 Pro30.57
  • GPT-5.4 Pro30
  • Claude Fable 528.57
  • GPT-5.527.14
  • GPT-5.423.43

Head-to-head

  • GPT-5.5 Pro vs GPT-5.4 Pro
  • GPT-5.5 Pro vs Claude Fable 5
  • GPT-5.4 Pro vs Claude Fable 5