The paper examines test‑time scaling (TTS) methods that use extra inference compute to improve language model outputs. Across five open‑ended benchmarks—medicine, law, finance, general chat, and creative writing—the study finds that increasing exploration (generating more candidates) consistently yields better top candidates, but exploitation (selecting the final output) remains weak due to poor reward‑model correlation. Only the Fusion approach, which synthesizes candidates, reliably improves results, yet it recovers only about 40% of the potential quality, indicating that the bottleneck lies in choosing from the candidate pool rather than generating it.
arXiv:2609.15309v1 Announce Type: new Abstract: Large language model (LLM) agents allocate test-time compute adaptively as they revise solutions, use tools, explore alternatives, and decide when to s...
The paper demonstrates that the number of candidates generated during test-time scaling of large language models does not fully capture the system cost. By comparing different generation schedules (e.g., one batched call versus multiple serial calls) while keeping the total candidate count fixed, the authors show that serial calls consume significantly more GPU energy and latency. The study suggests that reporting candidate count alone is insufficient; evaluations should also include generation schedule and GPU-level metrics.
The paper investigates how test‑time computation can enhance language models and at what cost, introducing the SELF‑POT benchmark to evaluate this across competition mathematics, competitive programming, and agentic workflows. SELF‑POT separates candidate coverage from final accuracy, tracks correctness transitions under revision, and measures protocol completion alongside task success. Using a unified budget rule, the study compares direct inference, parallel sampling, and self‑revision across five low‑cost reasoning models, revealing that selection rules and failure handling significantly influence gains and cost savings.
The paper investigates test‑time scaling for language models, noting that sequential methods often stall due to entering attractors—sets of answers that block further exploration. Across 27 experiments, over half of sequential trajectories hit an attractor within four iterations. Introducing a simple model‑mixing intervention reduces attractor incidence by 21.2 percentage points, expands solution coverage beyond a compute‑matched parallel baseline, and boosts recursive self‑aggregation accuracy by at least 2.2 percentage points.
arXiv:2605. 25143v2 Announce Type: replace Abstract: Test-time scaling improves language model reasoning by spending additional compute to explore multiple solution trajectories.
arXiv:2607. 09438v1 Announce Type: cross Abstract: Test-time scaling (TTS) reliably improves reasoning in large language models, but whether it transfers to small open vision-language models remains unclear.
arXiv:2510. 11686v2 Announce Type: replace-cross Abstract: Reinforcement learning (RL) promises to expand the capabilities of language models, but it is unclear if current RL techniques promote the discovery of novel behaviors, or simply sharpen those already present in the base model.
The paper introduces BudgetAPO, a single‑stage prompt optimizer designed for tight‑budget scenarios where only a few hundred model calls are feasible. BudgetAPO adapts the evaluation slice size to each task’s noise level, uses a fixed slice to turn accept/reject decisions into paired comparisons, and applies a reflective operator to jointly rewrite reasoning strategy and output format. Across seven benchmarks and five subject models, BudgetAPO outperforms existing methods, ranking first on every subject and requiring far fewer calls to achieve comparable or superior performance.
The paper investigates how iterative language‑model optimizers improve prompts and programs by evaluating candidates, exposing failing cases to a proposer model, and retaining revisions that enhance performance. By conducting 1,750 pre‑registered experiments across nine benchmarks, the authors disentangle the effects of parent inheritance (keeping a surviving parent) from feedback disclosure (providing failure summaries) under equal computational budgets. They find that greedy parent inheritance alone drives most optimization gains, outperforming seed‑reset strategies even when the latter receive comprehensive failure summaries or larger proposer models, and that the cost of proposer reflection is minimal compared to validation and minibatch gating.
arXiv:2606. 01667v1 Announce Type: new Abstract: Test-time scaling has become a major way to improve large language model reasoning, but its orchestration has remained designer-engineered: a fixed sample budget, a fixed refinement loop, a fixed scoring rule, or a fixed search policy decides how compute is spent, leaving the model in charge of solving but not of orchestration.
The paper introduces Stitching Noisy Diffusion Thoughts, a self‑consistency framework that transforms low‑cost diffusion‑sampled reasoning into a reusable pool of step‑level candidates. By sampling many diverse trajectories with a masked diffusion language model, scoring each intermediate step with a process reward model, and stitching the highest‑quality steps across trajectories, the method recomputes only the final answer, separating exploration from evaluation. Experiments on math reasoning benchmarks show that step‑level recombination improves accuracy by up to 23.8% on six tasks while reducing latency by up to 1.8× compared to traditional diffusion models and unified architectures.
arXiv:2609.38672v1 Announce Type: cross Abstract: Beam-search-based test-time methods provide an effective way to improve large language model (LLM) performance on long-horizon generation by pruning...