arXiv AI July 22, 2026 By Abdallah Khemais (ISITCOM, University of Sousse)

Cost Accounting for Reactive Computational Graphs: Exhaustive Sweeps, Sequential Mutation, and the Backward-Locality Gap

arXiv:2607. 18323v1 Announce Type: cross Abstract: Exhaustive site-by-site interventions on a neural network's computational graph -- activation-patching sweeps, circuit-discovery searches, systematic ablation studies -- mutate the graph at every candidate site, and their cost is dominated by recomputation after each mutation.

Summary generated by The Flow from the publisher's feed. The full article lives at arXiv AI.

benchmarks

arXiv AI

Jul 21

Exact Network Surgery: Functional Invariance and Gradient Plasticity in Reactive Computational Graphs

arXiv:2607. 16568v1 Announce Type: new Abstract: Function-preserving network growth techniques such as Net2Net and progressive stacking expand a model's capacity without destroying its learned function, but existing formulations either tolerate numerical perturbations or require a full rebuild of the training program.

By Abdallah Khemais (ISITCOM, University of Sousse)

arXiv AI

Jul 24

When Does Recurrence Become an Algorithm? Convergence Selection in Weight-Tied Looped Transformers

arXiv:2607. 20594v1 Announce Type: cross Abstract: When does a weight-tied looped transformer -- one block applied T times -- implement an actual algorithm?

By Tong Zhang, Junhao Hu, Yun Peng, Tao Xie

llms benchmarks

arXiv Machine Learning

Jul 7

ParEVO: Synthesizing Code for Irregular Data: High-Performance Parallelism through Agentic Evolution

arXiv:2603. 02510v2 Announce Type: replace Abstract: The transition from sequential to parallel computing is essential for modern high-performance applications but is hindered by the steep learning curve of concurrent programming.

By Liu Yang, Zeyu Nie, Andrew Liu, Felix Zou, Deniz Altinb\"uken, Amir Yazdanbakhsh, Quanquan C. Liu

llms agents fine-tuning benchmarks

arXiv Machine Learning

Jun 8

Reversible Foundations: Training a 120B Sparse MoE through State-Preserving Scaling

arXiv:2606. 07404v1 Announce Type: new Abstract: This paper reports on training a hundred-billion-parameter sparse mixture of experts on a single eight-GPU node, end to end.

By Rohan Shravan

llms nlp efficiency

arXiv AI

Jul 22

Circuit Claims Depend on What Is Extracted and How It Is Compared

arXiv:2607. 18921v1 Announce Type: cross Abstract: Circuit extraction identifies a small set of model components whose presence preserves a target behavior under ablation, and the resulting circuit is often read as the mechanism behind that behavior.

By Yang Sheng, Jie Fu

llms reinforcement-learning efficiency benchmarks

arXiv AI

Jun 9

Capacity-Controlled Global Attention for Graph Transformers

arXiv:2604. 17324v2 Announce Type: replace-cross Abstract: Global self-attention drives modern graph transformers, yet the softmax at its core imposes a structural constraint rarely examined directly: every attention row is non-negative and sums to one, so each per-head output is a mass-conserving convex combination of value vectors.

By Yang Liu, Dongxin Guo, Tom Zheng, Siu Ming Yiu, Liam Ning, Jikun Wu

llms benchmarks

Related stories

Exact Network Surgery: Functional Invariance and Gradient Plasticity in Reactive Computational Graphs

When Does Recurrence Become an Algorithm? Convergence Selection in Weight-Tied Looped Transformers

ParEVO: Synthesizing Code for Irregular Data: High-Performance Parallelism through Agentic Evolution

Reversible Foundations: Training a 120B Sparse MoE through State-Preserving Scaling

Circuit Claims Depend on What Is Extracted and How It Is Compared

Capacity-Controlled Global Attention for Graph Transformers