---
title: "RISC Beats Majority Voting Ceiling by 8% With a Tiny Ranker"
subtitle: "RISC replaces majority voting with a learned ranker that cuts LLM calls by up to 82% while also beating peak self-consistency accuracy on QA benchmarks"
url: "https://alphasignal.ai/news/risc-beats-majority-voting-ceiling-by-8-with-a-tiny-ranker"
author: "AlphaSignal Newsroom"
author_url: "https://alphasignal.ai/editorial-team"
company: "AlphaSignal"
company_url: "https://alphasignal.ai/company/alphasignal"
published: "2026-06-03T00:00:00.000Z"
updated: "2026-06-12T01:06:50.735Z"
type: "Paper"
topics:
  - "Reasoning"
  - "Benchmarks"
subtopics:
  - "Test Time Compute"
  - "Math Reasoning"
upvotes: 1660
source_url: "https://arxiv.org/abs/2606.05054"
comments: 0
truncated: true
---

# RISC Beats Majority Voting Ceiling by 8% With a Tiny Ranker

> RISC replaces majority voting with a learned ranker that cuts LLM calls by up to 82% while also beating peak self-consistency accuracy on QA benchmarks

Self-consistency is one of the simplest and most effective tricks in the LLM playbook: sample the same question multiple times, then pick the most common answer. Instead of trusting a single chain, the method selects the answer most consistent across multiple reasoning paths, with the intuition that different valid reasoning chains should converge on the same correct answer. It works surprisingly well. The problem is that it wastes a lot of what it generates.

A large gap remains between majority voting and an oracle that always picks the correct answer from the sampled candidates. In other words, the right answer is usually already in the pile. The model just can't reliably find it. A new paper from researchers at AIRI, Skoltech, and EPFL proposes a fix: stop voting, start ranking.

## The gap majority voting can't close

Several prior works have attempted to identify robust signals that help distinguish the correct answer among sampled candidates, including CISC, which relies on verbal uncertainty estimates, and methods based on internal geometric structure such as stable rank. These approaches all share the same fundamental design philosophy: find one good signal and use it. Such hand-crafted, knowledge-driven methods often struggle to scale and generalize to new tasks, a pattern long observed in AI research.

The new method, [Ranking-Improved Self-Consistency (RISC)](https://arxiv.org/abs/2606.05054), takes a different approach. Rather than picking a single signal and hoping it generalizes, it trains a small model to combine multiple signals together, treating answer selection as a ranking problem rather than a counting problem.

## Five signals, one lightweight ranker

The core of RISC is a [LightGBM LambdaRank model](https://github.com/s-nlp/RISC) \-- a gradient-boosted tree that learns to order candidates by correctness. LambdaRank is a classic learning-to-rank algorithm that optimizes directly for ranking quality (NDCG) rather than raw classification accuracy. The key insight is that it scores candidates **relative to each other**, not independently, which lets it exploit agreement structure across the full sample pool.

_Preview only. Full text requires an AlphaSignal Pro account._

## Sources

- [Boosting Self-Consistency with Ranking (arXiv)](https://arxiv.org/abs/2606.05054)
- [RISC Paper HTML](https://arxiv.org/html/2606.05054)
- [RISC GitHub Repository](https://github.com/s-nlp/RISC)
- [LLM Post-Training: A Deep Dive into Reasoning Large Language Models](https://arxiv.org/pdf/2502.21321)
- [Confidence-Weighted Token Set Cover for Early Hypothesis Pruning in Self-Consistency](https://arxiv.org/pdf/2508.03979)