CAPS is a four-stage inference-only cascade that adapts how much of each solution the verifier sees and how comparisons are distributed, halving per-candidate verifier tokens while outperforming uniform pairwise verification on most benchmarks.
arXiv preprint arXiv:2509.26626 , year=
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
years
2026 5verdicts
UNVERDICTED 5representative citing papers
EvoLib enables LLMs to accumulate, reuse, and evolve knowledge abstractions from inference trajectories at test time, yielding substantial gains on math reasoning, code generation, and agentic benchmarks without parameter updates or supervision.
ZAYA1-8B is a reasoning MoE model with 700M active parameters that matches larger models on math and coding benchmarks and reaches 91.9% on AIME'25 via Markovian RSA test-time compute.
A 4B model post-trained with SFT, RL, and a reasoning cache surpasses larger open models and approaches proprietary ones on Olympiad proof generation.
Lack of exploration from conditioning on prior answers is the primary reason parallel sampling outperforms sequential sampling in large reasoning models.
citing papers explorer
-
CAPS: Cascaded Adaptive Pairwise Selection for Efficient Parallel Reasoning
CAPS is a four-stage inference-only cascade that adapts how much of each solution the verifier sees and how comparisons are distributed, halving per-candidate verifier tokens while outperforming uniform pairwise verification on most benchmarks.
-
ZAYA1-8B Technical Report
ZAYA1-8B is a reasoning MoE model with 700M active parameters that matches larger models on math and coding benchmarks and reaches 91.9% on AIME'25 via Markovian RSA test-time compute.
-
QED-Nano: Teaching a Tiny Model to Prove Hard Theorems
A 4B model post-trained with SFT, RL, and a reasoning cache surpasses larger open models and approaches proprietary ones on Olympiad proof generation.