arXiv Machine Learning

Training, learning and inference: unified dynamics of neural systems

arXiv Machine Learning
Nov 6

Unified dynamics of training, learning, inference, and reinforcement in neural networks

The paper reports nine nanoGPT experiments that uncover relationships among training, learning, inference, and reinforcement in neural networks. It shows that updates trigger state‑dependent responses, learning reorganizes functional support, and inference uses this support without altering it. A predictor derived from the fifth experiment accurately forecasts post‑update capability transitions, and the authors extend these findings to other architectures, demonstrating that balanced feedback can restore lost skills and form a unified dynamical framework for neural‑network behavior.

By Mian Wang
arXiv Machine Learning
Sep 25

Intrinsic-Extrinsic Coupling in Learning Dynamics

The paper introduces a framework for intrinsic‑extrinsic coupling in learning dynamics, defining it via a continuation‑conditioned value of a constrained learning‑state intervention and observation‑relative fibers. It presents an executable finite‑frame classifier‑head that protects current logits while repairing historical margins, and distinguishes local admissibility, intervention value, and complete‑policy performance. Experiments on CLINC‑derived class‑incremental tasks, output distillation with RoBERTa, and SGDW dynamics demonstrate that coupling can produce both positive and negative interactions, and that coordinated content controls can match or exceed development gains while guided allocation reduces cross‑entropy loss compared to standard replay.

By Qinyou Wang
Hugging Face Trending Papers
Sep 24

Intrinsic-Extrinsic Coupling in Learning Dynamics

The paper introduces a framework for intrinsic‑extrinsic coupling in learning dynamics, where a learner’s current observations do not solely dictate its future training responses. It formalizes this coupling through a continuation‑conditioned value of a constrained learning‑state intervention and employs an executable finite‑frame classifier‑head to protect current logits while adjusting historical margins. Experiments across CLINC‑derived class‑incremental settings, output distillation with RoBERTa, and SGDW dynamics reveal that coupling can produce both positive and negative interactions, and that coordinated interventions can match or exceed development gains while reducing cross‑entropy loss compared to standard replay.

arXiv AI
Sep 4

FlowBalance: Verifier-Grounded Self-Improvement from On-Policy Reasoning Experience

arXiv:2609. 03241v1 Announce Type: cross Abstract: A reasoning model can improve from its own on-policy experience, but this inner loop is fragile: terminal verifiers provide reliable yet sparse supervision, while dense same-model guidance can reinforce false confidence or overconcentrate learning on a narrow solution mode.

By Zixun Huang, Kishan Panaganti, Haitao Mi, Leowei Liang
arXiv Machine Learning
Aug 18

Metaplasticity as adaptive gradient preconditioning for incremental learning

arXiv:2608. 14634v1 Announce Type: new Abstract: Biological intelligence naturally prevents catastrophic forgetting through Complementary Learning Systems (CLS) theory, a macroscopic consolidation process driven at the local level by synaptic metaplasticity: the continuous, history-dependent neuromodulation of individual synapses.

By Isabelle Aguilar, Zayn Andre Zainal, Omid Kavehei
arXiv Computation and Language
Sep 28

Recursive Self-Improvement via On-Policy Distillation for Reasoning

The paper introduces a recursive self-improvement framework for language models that replaces an external teacher with a frozen copy of the student, enabling dynamic co-evolution (DCE) and self-refined concise learning (SRCL). DCE allows the privileged teacher to evolve alongside the student, while SRCL trains on shorter, verified rewrites to reduce verbosity. Experiments show that the combined DCE+SRCL approach outperforms traditional on‑policy self‑distillation across multiple model sizes and math benchmarks, achieving significant accuracy gains and shorter outputs.

By Shangjian Yin, Zehao Zhao, Kavosh Asadi, Rui Liu, Yuchen Lu, Shike Mei, Hang Cui, Luke Simon, Zhouxing Shi, Hamed Firooz