Entity · model

Fixed-Point Reasoning Model

modelactivefixed-point-reasoning-model-8e50321e·1 events·first seen Jun 17, 2026

Aliases: Fixed-Point Reasoning Model

Co-occurring entities

Fixed-Point Reasoners: Stable and Adaptive Deep Looped Transformers ARC-AGI

More like this (12)

Fixed-Point Reasoners: Stable and Adaptive Deep Looped Transformers Predicting Future Behaviors in Reasoning Models Enables Better Steering Reasoning Language Models When the Chain of Thought Knows Better: Failure Modes in Multi-Turn Reasoning Models frontier reasoning models Large Reasoning Models Reasoning in Memory (RiM)mathematical reasoning Does Reasoning Preserve Alignment? On the Trustworthiness of Large Reasoning Models Adaptive Parallel Reasoning Beyond the Commitment Boundary: Probing Epiphenomenal Chain-of-Thought in Large Reasoning Models Equilibrium Reasoners (EqR)

Recent events (1)

5arXiv · cs.AI·Jun 17, 2026·source ↗

Fixed-Point Reasoning Model (FPRM): Stable looped Transformers with adaptive compute via fixed-point halting

Researchers introduce FPRM, a Transformer-based Fixed-Point Reasoning Model that uses fixed-point convergence as a halting mechanism in looped architectures, addressing signal propagation problems through pre-norm layers and residual scaling. Looped architectures provide inductive bias for compositional reasoning, but suffer from depth-induced signal degradation when halting is deferred; FPRM resolves this while enabling compute to scale with task difficulty. The model is evaluated on Sudoku, Maze, state-tracking, and ARC-AGI benchmarks. This contributes to the growing body of work on adaptive-compute and iterative-refinement architectures for reasoning.

Evaluation and Benchmarking Fixed-Point Reasoning Model Fixed-Point Reasoners: Stable and Adaptive Deep Looped Transformers ARC-AGI