---
title: "Value-Conditioned Proposal Acquisition for Test-Time Reasoning"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2611.02852v1"
permalink: "https://feed.astralor.com/items/itm_a0fff9014d77d4e939b8ef1e"
published: "2026-11-02T09:53:39.000Z"
score: 57
channel: "finance"
tags: ["RAG","Value-Conditioned Proposal Acquisition","Test-Time Reasoning Sampling","ValueFormation","Sampling","MATH-500","Llama-fitted"]
---

# Value-Conditioned Proposal Acquisition for Test-Time Reasoning

## At a glance

Sampling and selection spend test-time compute inside a candidate pool, but they cannot recover a correct answer that never enters it. We introduce ValueFormation, a test-time policy that decides after each primary-mode…

## Key points

- Sampling and selection spend test-time compute inside a candidate pool, but they cannot recover a correct answer that never enters it.
- We introduce ValueFormation, a test-time policy that decides after each primary-model draw whether to stop, continue sampling, or acquire one proposal from a heterogeneous reasoning specialist.
- The three actions are compared on one cost-regularized value scale learned from vote-state features; the calibration is frozen across tasks and primary models, and acquired proposals are accepted only through a frozen p…

## Full text

Sampling and selection spend test-time compute inside a candidate pool, but they cannot recover a correct answer that never enters it. We introduce ValueFormation, a test-time policy that decides after each primary-model draw whether to stop, continue sampling, or acquire one proposal from a heterogeneous reasoning specialist. The three actions are compared on one cost-regularized value scale learned from vote-state features; the calibration is frozen across tasks and primary models, and acquired proposals are accepted only through a frozen posterior comparison. On held-out MATH-500, one Llama-fitted calibration improves over the strongest of eight identically evaluated baselines by more than fifteen points for each of three 8B primaries and transfers unchanged to Qwen3-14B. Removing acquisition lowers Llama MATH accuracy by 20.43 points, while same-trace and same-access replays attribute the gain to state-conditioned acquisition rather than specialist access alone. Where primary coverage is high, as on GSM8K, acquisition is rare and gains mainly come from cheaper allocation; on MBPP, a fixed code route attains the top accuracy at the lowest interaction cost. Varying the token price monotonically reduces calls without retraining.

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_a0fff9014d77d4e939b8ef1e