Benchmarks · composite
Vals RSI Index v1.1: Pre-training: LM Training
Can a model do the research that builds the next model? Autonomous LLM R&D scored against frozen published, human, and frontier-model references. Published Pre-training: LM Training score in percentage points.
21 models · 21 results · updated 2026-10-15 · vals.ai ↗
Leaderboard
Opus 5.5 max
More from this benchmark
6 published measuresScore vs any published measure6 measures: sub-scores, splits, counts
Breakdown
darker is better within each column · click a column to sort| Model | |||||
|---|---|---|---|---|---|
| Gemini 4 Argon max | 59.8M | 1.1M | 516k | 36.2M | 0 |
| GPT-6.1 Sol max | 222.2M | 1.5M | 1.2M | 213.5M | 8.3M |
| Grok 4.7 max | 137.3M | 833k | 551k | 119.2M | 0 |
| GPT-6 Sol max | 1189.5M | 872k | 465k | 1180.1M | 9.4M |
All results
21 rows| # | Model | Effort | Score |
|---|---|---|---|
| 1 | Opus 5.5 | max | 53.0% |
| 2 | Fable 5.1 | max | 49.5% |
| 3 | GPT-6 Astra | max | 48.4% |
| 4 | Gemini 4 Argon | max | 48.0% |
| 5 | GPT-6.1 Sol | max | 42.9% |
| 6 | Opus 5 | max | 42.8% |
| 7 | Grok 4.6 | max | 42.3% |
| 8 | Grok 4.7 | max | 41.6% |
| 9 | Muse Spark 1.3 | max | 40.9% |
| 10 | Fable 5 | max | 40.8% |
| 11 | Opus 4.8 | max | 40.8% |
| 12 | GLM 5.3 | max | 40.8% |
| 13 | Kimi K3 | max | 40.4% |
| 14 | Gemini 3.7 Flash | max | 39.7% |
| 15 | Opus 4.7 | max | 39.3% |
| 16 | Qwen 3.8 Max | max | 38.3% |
| 17 | Gemini 3.8 Flash | max | 36.2% |
| 18 | GPT-5.6 Sol | max | 36.1% |
| 19 | GPT-5.5 | xhigh | 31.1% |
| 20 | GPT-6 Sol | max | 31.1% |
| 21 | GPT-5.4 | xhigh | 28.0% |