# Researchers find Qwen3’s reasoning gains may stem from ‘thinking leakage’

Digest AI · Research · published 2026-10-14T04:00:00Z

Canonical: https://digestai.news/story/researchers-find-qwen3s-reasoning-gains-may-stem-from-thinking-leakage

## Summary

A new paper on arXiv examines how reinforcement learning (RL) training affects hybrid reasoning models like Qwen3. The study suggests that much of the reported improvement in reasoning performance—when models switch between THINK and NO-THINK modes—actually comes from *thinking leakage*: the model partially reactivates its original THINK mode rather than learning new reasoning behavior.

The authors trained Qwen3 variants (1.7B, 4B, and 8B parameters) using GRPO in NO-THINK mode. As accuracy rose, responses grew longer and more reflective, resembling the original model’s THINK mode. Weight and activation analysis showed that further RL scaling merely shifted behavior toward THINK mode without exceeding its baseline accuracy. Swapping activation subspaces between RL-trained and original models confirmed this: the original model absorbed most of the RL gain when given its THINK-related activations.

## Key points

- Qwen3’s reasoning gains in NO-THINK mode may reflect reactivation of original THINK mode, not new learning
- RL-trained models’ responses grow longer and more reflective, mirroring THINK mode behavior
- Swapping activation subspaces shows original model absorbs most RL gains when given THINK-related data

## Why it matters

If true, this challenges assumptions about RL’s role in improving reasoning—suggesting gains may be superficial rather than transformative.

## Sources

1. [The Thinking Leakage in Reinforcement Learning for LLMs](https://arxiv.org/abs/2610.14270) (arXiv cs.CL, 2026-10-14, primary source)

## Cite

Digest AI, "Researchers find Qwen3’s reasoning gains may stem from ‘thinking leakage’", 14 October 2026, https://digestai.news/story/researchers-find-qwen3s-reasoning-gains-may-stem-from-thinking-leakage

---

Written by Digest AI's editorial model from the linked sources; the sources are the record. Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse
JSON: https://digestai.news/story/researchers-find-qwen3s-reasoning-gains-may-stem-from-thinking-leakage.json