LLM benchmark gains often come from producing already-reachable answers more reliably, not from making new answers reachable under a matched probe.
Router-Tuning: A Simple and Effective Approach for Dynamic Depth
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Reachability Is Not Realization: Tracing the Sources of LLM Benchmark Gains
LLM benchmark gains often come from producing already-reachable answers more reliably, not from making new answers reachable under a matched probe.