GPT-4 Turbo 04/09 · Michelangelo (Latent Structure Queries) · MRCR · cumulative mean through 128K · chart reading · Michelangelo original report · MRCR at 128K context ↗
OpenAI / MODEL EVIDENCE
GPT-4 Turbo 04/09
1 benchmarks · 1 observations · 1 sources
Scores at a glance
1 benchmarks · large dot: GPT-4 Turbo 04/09; small dots: closest models where reported
Closest models by score
Similar positions on shared benchmark leaderboards. Smaller gaps are closer.
GPT-4o →OpenAI
1 of 1 shared benchmark comparisons
Claude 3 Opus →Anthropic
1 of 1 shared benchmark comparisons
Gemini 1.5 Flash 08/27 →Google
1 of 1 shared benchmark comparisons
How similarity is measured
Mean absolute gap in percentile positions on shared benchmark settings; lower is closer. At least 60% overlap is required, with a minimum of three when available. Independent evaluations are used when there are at least three; otherwise published report comparisons are used. Repeated model observations within one setting are averaged before ranking. Writing-style traits and composite indices are excluded. This describes similarity, not an overall quality ranking.
Sources
Every plotted observation links to its evaluation. Reported configurations, benchmark versions and task subsets can differ; similarity describes shared scores, not training lift.
- Michelangelo: Long Context Evaluations Beyond Haystacks via Latent Structure Queries — published result table ↗
All 24 authors
Vodrahalli, Kiran, Ontanon, Santiago, Tripuraneni, Nilesh, Xu, Kelvin, Jain, Sanil, Shivanna, Rakesh, Hui, Jeffrey, Dikkala, Nishanth, Kazemi, Mehran, Fatemi, Bahare, Anil, Rohan, Dyer, Ethan, Shakeri, Siamak, Vij, Roopali, Mehta, Harsh, Ramasesh, Vinay, Le, Quoc, Chi, Ed, Lu, Yifeng, Firat, Orhan, Lazaridou, Angeliki, Lespiau, Jean-Baptiste, Attaluri, Nithya, Olszewska, Kate
Names used in source tables
GPT-4 Turbo 04/09