CL-bench Life
Integrity rank #41 of 65 · 13 models scored · top score 22.2 · GPT-5.5
Context learning over messy, real-life, unstructured everyday context
Strongest on contamination resistance, weakest on discrimination. Caveat: the public test set does not rule out training-data contamination.
Reported scores — protocols may differ. How we compare
| # | Model | Score | Evidence | Measured |
|---|---|---|---|---|
| 1 | GPT-5.5 | 22.2 | unverifiedT2 | 2026-04-23 |
| 2 | GPT-5.4 | 21.7 | unverifiedT2 | 2026-03-05 |
| 3 | GPT-5.1 | 17.3 | unverifiedT2 | 2025-11-13 |
| 4 | Claude Opus 4.6 | 17 | unverifiedT2 | 2026-02-05 |
| 5 | Gemini 3.1 Pro | 16.9 | unverifiedT2 | 2026-02-19 |
| 6 | DeepSeek-V4-Pro | 13.5 | unverifiedT2 | 2026-04-24 |
| 7 | Kimi K2.5 | 13.2 | unverifiedT2 | 2026-01-27 |
| 8 | Qwen 3.5 Plus (hosted 397B-A17B) | 12.4 | unverifiedT2 | 2026-02-16 |
| 9 | Grok 4.20 | 11.9 | unverifiedT2 | 2026-02-17 |
| 10 | GLM-4.7 | 10.9 | unverifiedT2 | 2025-12-22 |
| 11 | DeepSeek-V3.2-Exp | 9.5 | unverifiedT2 | 2025-09-29 |
| 12 | DeepSeek-V3.2 | 7.4 | unverifiedT2 | 2025-12-01 |
| 13 | MiniMax-M2.5 | 6.3 | unverifiedT2 | 2026-02-12 |
Continue from this leaderboard into the tightest head-to-head reads.
Weighted composite. Each dimension opens to the source-backed sub-signals it is computed from.
Claims drawn from cited facts, not live model generation.
This benchmark scores 13 models, with a spread of 15.9 points between the best and worst results. That wide gap supports treating the rank differences between models as a real separation in capability rather than noise from close scores.
3 cited facts
This benchmark registers an integrity score of 81, graded B, ranking 39th of 62 benchmarks in the universe. That score reflects the benchmark's health as a discriminator of frontier models under a disclosed harness, not any model's capability. Its weakest component is discrimination: with frontier scores bunched together, the benchmark separates leading models less sharply than its other checks would allow.
5 cited facts
This benchmark is not saturated: the top score is 22.2 and there are 77.8 points of headroom left to the ceiling, with two models clustered within a couple of points near the top. That means there is still genuine room to separate the field at the top, so the current ranking is not merely noise among leaders.
5 cited facts
Because this benchmark uses a consistent harness, scores are directly comparable within it, yet the public test set means high task performance under the harness warrants more scrutiny for possible contamination than a held-out set would. This benchmark's contamination history includes bespoke everyday-context scenarios designed to resist contamination. Its rubric-based grading with per-category solving rates carries LLM-judge-sensitivity caveats, so scores should be read as task performance under the disclosed harness rather than deployed capability.
4 cited facts
This benchmark is designed to measure context learning over messy, real-life, unstructured everyday context. The test was built as a set of expert-curated context-task pairs with multiple verification rubrics across several categories, including communication and social interactions, fragmented information and revisions, and behavioral records and activity trails. Its design assumes that messy real-life contexts paired with rubric grading can proxy real-world context-learning ability, and that its categories represent everyday information settings. The sharpest caveat is that it is small, uses strict all-rubric scoring, may suffer from rubric and judge noise, and its scenarios are curated rather than naturally sampled; no human baseline is documented.
5 cited facts
CL-bench Life: CL-bench Life is a companion to CL-bench that shifts from clean professional sources to messy real-life context - everyday communication, fragmented notes, and behavioral traces that are socially grounded and temporally dispersed - testing whether models can learn from and reason over unstructured everyday information.
GPT-5.5 leads CL-bench Life at 22.2. The full leaderboard above lists every recorded measurement, not just the headline number.
13 models have recorded scores on CL-bench Life, spanning a score spread of 15.9.
tensor.news grades CL-bench Life B for integrity (score 81/100), ranking #41 of 65 benchmarks we assess across discrimination, saturation, contamination resistance, harness comparability, and freshness.
No — CL-bench Life still has headroom and continues to discriminate between models rather than bunching them at the ceiling.
Its test set is public, so contamination risk is on the table: a high score may partly reflect training-data overlap rather than capability.
Scores are reported under a consistent harness, so comparisons on CL-bench Life are reasonably apples-to-apples.
Every CL-bench Life measurement is source-backed and tagged reproduced, self-reported, or unverified; the underlying record cites arxiv.org/abs/2604.27043.