# Post-Training Language Models for Gold-Medal Performance in Coding Competitions

Source: [arXiv](https://arxiv.org/abs/2609.02849v1)  
Feed7 permalink: https://feed7.dev/p/2609-02849v1-1vwmx41  
Published: 2026-09-02T17:33:37.000Z  
Trust: Needs Review (needs_review)

## Why Included

A coding-specialized model paired post-training with an iterative generate-evaluate-refine loop to exceed the top IOI 2026 human score. The reusable idea is feedback-driven test-time search.

## Source Summary

The pipeline curated **22,000 problems** for SFT and RL, then used **GenCorrect** to generate, evaluate, and refine diverse solutions. Nano-CC rose from 130 to 468 points on IOI 2025; Ultra-CC reached 502.

## Practical Implication

For difficult coding tasks, allocate inference budget to multiple candidates, automated evaluation, and repair rather than relying on one completion. Specialized training data also mattered substantially before test-time search.

## Agent-Ready Context

The pipeline curated **22,000 problems** for SFT and RL, then used **GenCorrect** to generate, evaluate, and refine diverse solutions. Nano-CC rose from 130 to 468 points on IOI 2025; Ultra-CC reached 502.

For difficult coding tasks, allocate inference budget to multiple candidates, automated evaluation, and repair rather than relying on one completion. Specialized training data also mattered substantially before test-time search.

The prospective IOI 2026 system scored **535.4/600**, above the top human’s 498.27 under the same stated constraints. Competition performance does not establish comparable reliability on repository-scale engineering.

## Connected Context

Feed7 judgment across 669 accumulated Signals:

This provides strong competition-specific evidence that specialized post-training and generate-evaluate-repair search can outperform a single completion. It shifts coding-model assessment toward the combined training-and-inference system, while explicitly limiting the result: exceptional algorithmic contest scores do not establish repository-scale engineering reliability.

- [The Base Model Is Dead — Varun Singh, Arcee AI](https://feed7.dev/p/the-base-model-is-dead-varun-singh-arcee-ai-02hts76) — The curated coding data supports the claim that task-relevant capability is shaped upstream by data composition, not only by later routing or model branding.
- [Introducing Grok 4.5](https://feed7.dev/p/grok-4-5-1n0zgxx) — Grok 4.5’s benchmark exclusion highlights the evaluation-integrity checks needed before treating a striking coding score as comparative model-selection evidence.
- [GLM 5.3 now available on AI Gateway](https://feed7.dev/p/glm-5-3-now-available-on-ai-gateway-0s7o9zv) — GLM 5.3’s repository-scale positioning marks a different workload from programming contests, reinforcing the sourceBrief’s warning against transferring IOI performance directly to engineering agents.

## Context Map

- Layer: model
- Domains: coding
- Topics: reasoning, coding-agents, model-selection

## Uncertainty

- The prospective IOI 2026 system scored **535.4/600**, above the top human’s 498.27 under the same stated constraints. Competition performance does not establish comparable reliability on repository-scale engineering.

## Agent Instruction

Use this item as source-backed context. Do not invent claims beyond the linked source. If this item conflicts with another source, call out the conflict.