Google Research turns model routing into measurable diversity and stability metrics
burkov · x · 2026-07-21
- A new Google Research paper studies how routers choose among multiple language models or agents.
- The authors argue that accuracy and cost alone miss two issues: some models behave too similarly for routing to matter, and tiny wording changes can send the same request to different models.
- They define two measurable quantities: one for behavioral diversity across models, and one for routing stability under meaning-preserving edits such as typos, paraphrases, or irrelevant additions.
- Experiments suggest that a carefully selected set of fewer than ten models can preserve most useful diversity from much larger pools.
- Nearest-neighbor routers can be accurate yet brittle, while description-based routers may be less accurate in some cases but more stable.
Related event: Google DeepMind Proposes Metrics for Model Routing Stability(2 posts)→
More from Models
- BullshitBench update: GPT-6-Astra beats all prior OpenAI models but still trails Anthropic — scaling01 · 2026-09-11
- Astra Scores 83% on GauntletBench, First Computer-Use Agent to Beat Human Baseline — ducha_aiki · 2026-09-11
- Kimi K2.8 Preview rolls out: near-K3 coding performance, 1M context for all tiers — teortaxesTex · 2026-09-11
- Looking for a classifier of software engineering task shapes to pick models per task — StewartalsopIII · 2026-09-11
- DeepSeek V4 Pro API to continue after Sept 2026, billing unchanged — teortaxesTex · 2026-09-11
- DeepSeek V4.1 Flash Hits 98% of GPT-6 Astra's Score at 1.4% of the Cost in Third-Party Benchmark — ayushtweetshere · 2026-09-11