A 3B Model Just Matched DeepSeek V3.2 on Math (671B Parameters, 223x Larger)

A 3-billion-parameter model has matched the performance of DeepSeek V3.2, a 671-billion-parameter model, on math benchmarks, challenging the assumption that larger models are always better.

Every reasoning model leaderboard tells the same story right now: bigger wins. Continue reading on Towards AI »