Unverified50% confidenceFactExact time
Model scores on competition mathematics (such as AIME and IMO) and code generation (such as Codeforces competition problems) have shown a near-vertical growth curve over the past year.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/9/2026
First Seen
Valid until: 1/7/2027
Sources
Related Entities
Related Claims
UnverifiedModel performance in non-verifiable domains has been steadily improving, but at a pace far slower than in mathematics and code.64% similarUnverifiedStarting this year, the growth rate of inference compute has already surpassed that of training compute.64% similarUnverifiedWhen a model improves significantly in the dimensions of math and code, averaging those gains with other dimensions that have remained stagnant makes the overall curve appear unremarkable.63% similarUnverifiedOn six new International Mathematical Olympiad problems, all models with room for improvement scored higher than their own providers' coding agents under the AutoFyn framework.63% similarUnverifiedIn code capability benchmarks such as SWE-bench, the performance gap for the same underlying model under different harnesses can reach tens of percentage points62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/991125API
curl https://kongchang.com/api/v1/knowledge/claims/991125MCP
get_claim(id=991125)