# Mazes

Mazes are modeled and solved in C++ using abstract data types, grids, paths, and graph-related algorithms.

This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles.

## Simpler Tiny Recursive Models

DevFeed: [Simpler Tiny Recursive Models](<https://devfeed.tech/articles/simpler-tiny-recursive-models-143431.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.23677>)

Author: Torben Glass, Zhengxi Zhang, Paul Swoboda

Published: 2026-11-19T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [llm-reasoning](<https://devfeed.tech/topics/llm-reasoning.md>), [machine learning overfitting](<https://devfeed.tech/topics/machine-learning-overfitting.md>), [Mazes](<https://devfeed.tech/topics/maze.md>)

Tags: [agi](<https://devfeed.tech/tags/agi.md>), [arc](<https://devfeed.tech/tags/arc.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [deep](<https://devfeed.tech/tags/deep.md>), [nested](<https://devfeed.tech/tags/nested.md>), [performance](<https://devfeed.tech/tags/performance.md>), [puzzle](<https://devfeed.tech/tags/puzzle.md>), [recursion](<https://devfeed.tech/tags/recursion.md>), [speed](<https://devfeed.tech/tags/speed.md>), [tiny](<https://devfeed.tech/tags/tiny.md>), [training](<https://devfeed.tech/tags/training.md>)

### AI overview

STRM simplifies Tiny Recursive Models (TRM) to a single state with a uniform update, using exponential moving updates at varied, training-jittered rates and reading solutions only from slow-updating channels. The abstract reports faster training and better final performance without sacrificing generalization. Under the same data, parameters and compute as TRM, STRM scores 3.8 points higher on Sudoku-Extreme and 8.4 higher on PPBench, and matches it on Maze-Hard. Against published TRM results, it scores 5.4 points higher on ARC-AGI-1 and 0.5 higher on ARC-AGI-2. These claims reflect the reported benchmarks; the supplied text is an abstract, not the full paper.

### Source excerpt

arXiv:2611.23677v1 Announce Type: new Abstract: Small models that execute deep looped computational graphs have shown great experimental results on puzzle-like problems like Sudoku, finding shortest paths through mazes, etc. The most well-known one is TRM (tiny recursive model), a simplification of HRM (hierarchical reasoning model). However, the underlying recursion is complicated and involves a high-level and a low-level state, a nested loop recursion, and the network calls with different input and output signatures depending on where we are in the loops. These mechanisms are needed to avoid memorization and overfitting. We propose STRM, a simplified design with a single state and a uniform update. STRM does not sacrifice generalization, improves training speed and final performance. Our key architectural choices that allow our simpler recursion to work are exponential moving updates on the states with varying rates that are jittered during training and readout of the solution on slow-updating channels only. We test STRM on Sudoku-Extreme, PPBench and Maze-Hard with the same data, parameters and compute as TRM: we improve over TRM by $3.8$ points on Sudoku-Extreme and $8.4$ points on PPBench and match it on Maze-Hard. On ARC-AGI-1 and ARC-AGI-2, STRM improves over the published TRM results by $5.4$ and $0.5$ points. We will release our code upon acceptance of the paper.

## Squingle Arcade Is Available Now For Free On PC VR

DevFeed: [Squingle Arcade Is Available Now For Free On PC VR](<https://devfeed.tech/articles/squingle-arcade-is-available-now-for-free-on-pc-vr-17308.md>)

Original publisher: [Read original article](<https://www.uploadvr.com/squingle-arcade-available-now-on-pcvr/>)

Author: Mike Johnson

Published: 2026-09-10T00:14:39Z

Content type: news

Language: en

Sources: [UploadVR](<https://devfeed.tech/sources/uploadvr.md>)

Topics: [arcade](<https://devfeed.tech/topics/arcade.md>), [Mazes](<https://devfeed.tech/topics/maze.md>), [pc](<https://devfeed.tech/topics/pc.md>)

Tags: [free](<https://devfeed.tech/tags/free.md>), [leaderboards](<https://devfeed.tech/tags/leaderboards.md>), [multiplayer](<https://devfeed.tech/tags/multiplayer.md>), [orbs](<https://devfeed.tech/tags/orbs.md>), [pc](<https://devfeed.tech/tags/pc.md>), [puzzle](<https://devfeed.tech/tags/puzzle.md>), [vr-gaming](<https://devfeed.tech/tags/vr-gaming.md>)

### AI overview

Squingle Arcade, a free-to-play quickplay version of the fluid-maze puzzle game Squingle, is now available on SteamVR and Quest. It includes more than 60 levels, asynchronous multiplayer, ghost racing, leaderboards, and scalable mazes.

### Source excerpt

Squingle Arcade, the follow up to the original fluid-maze-navigating puzzle game Squingle, is out now for PC VR.