FPRM is a Transformer-based model using fixed-point convergence for adaptive halting in looped architectures, claimed effective on Sudoku, Maze, state-tracking, and ARC-AGI benchmarks.
hub
Learning long-term dependencies with gradient descent is difficult.IEEE Transactions on Neural Networks, 5(2):157–166
17 Pith papers cite this work, alongside 6,592 external citations. Polarity classification is still indexing.
17
Pith papers citing it
6,592
external citations · Crossref
hub tools
citation-role summary
background 4
citation-polarity summary
roles
background 3polarities
background 3representative citing papers
Mamba-Assisted Closure (MAC) trains a Mamba sequence model on resolved trajectories to predict non-Markovian closures and couples it with reduced-order equations, outperforming Markovian, GRU, and Wilks baselines on Burgers' and Lorenz '96 systems.
A 1B-parameter hierarchical recurrent model pretrained on 40B instruction-response tokens achieves 60.7% MMLU and strong results on ARC-C, DROP, GSM8K, and MATH while using 100-900x fewer tokens than standard baselines.
MinMax RNCs are recurrent networks over the min-max semiring that achieve regular language expressivity, log-depth parallel scan, uniformly bounded states, and non-vanishing state gradients while showing competitive empirical performance.
Neural operators progressively forget domain geometry with depth due to Markovian layers and global mixing; a geometry memory injection mechanism mitigates this forgetting.
OGPO enables sample-efficient full-finetuning of generative control policies via off-policy critics and modified PPO, achieving SOTA on robot manipulation tasks while rescuing poorly initialized behavior cloning policies without expert data.
M²RNN achieves perfect state tracking at unseen lengths and outperforms Gated DeltaNet hybrids by 0.4-0.5 perplexity on 7B models with 3x smaller recurrent states.
mGRADE uses learnable-spaced convolutions shown to be equivalent to delay embeddings plus a lightweight gated recurrent component to achieve low-memory multi-timescale sequence modeling.
ModernBERT-labeled analysis of 306k social posts yields six major public-sentiment clusters on Advanced Air Mobility spanning noise, safety, regulation, workforce, drones, and military use.
BMRUs enable analog recurrent neural network hardware via discrete outputs that suppress noise 20-fold, with one-to-one parameter-to-circuit mapping and linear power scaling for recurrence.
CMRU restores gradient flow in BMRU via cumulative state updates with skip-connections through time, yielding better convergence and benchmark performance while retaining quantized persistent memory.
The serial scaling hypothesis formalizes inherently serial problems in complexity theory and demonstrates that diffusion models cannot solve them.
Systematic review of 145 papers on LLM-based log analysis, providing a unified taxonomy, common design patterns, evaluation practices, and challenges for deployment under drift and limited labels.
Established mathematical bottlenecks in representation, optimization, complexity, and high-dimensional learning aligned with the central disappointments of early AI research periods.
A tutorial framing deep learning as a complement to optimization for sequential decision-making under uncertainty, with applications in supply chains, healthcare, and energy.
citing papers explorer
-
Fixed-Point Reasoners: Stable and Adaptive Deep Looped Transformers
FPRM is a Transformer-based model using fixed-point convergence for adaptive halting in looped architectures, claimed effective on Sudoku, Maze, state-tracking, and ARC-AGI benchmarks.