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.
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
A learning system can occupy execution states that are indistinguishable under every declared present-behavior readout yet respond differently to future training. We formalize this through fiber fingerprints: controlled future-learning response laws restricted to present-behavior equivalence classes.
arXiv:2610.02700v1 Announce Type: new
Abstract: On-policy self-distillation (OPSD) supplies dense token-level feedback on trajectories sampled from the student's own policy, a richer training signal...
By Rui Li, Liyang He, Zheng Zhang, Zhenya Huang, Linbo Zhu, Qi Liu
arXiv:2608. 15976v1 Announce Type: new Abstract: A learning system can occupy execution states that are indistinguishable under every declared present-behavior readout yet respond differently to future training.
By Qinyou Wang
The paper presents a finite‑model synthesis that integrates operational state abstraction with optimal control within a stability‑evidence‑revision (SER) framework, termed "Poincaré meets Bellman." It distinguishes qualitative dynamics for reusable action‑response structures from dynamic programming that governs acquisition, retention, reuse, merging, and forgetting, and introduces a Bellman recursion over the joint law of hidden state and deployed memory. The authors derive explicit retention rules, demonstrate how factor sharing and informative observations improve identification, and verify coding and retention calculations through finite enumerations.
By Xin Li
arXiv:2608.20965v1 Announce Type: new
Abstract: We define an atomic generation fact f=(u,tau,omega,z;rho), recording the origin, realized transformation, concrete occurrence, generated result and rel...
By Mian Wang
The paper demonstrates that functional compatibility—how well new learning can coexist with behavior that must be preserved—is a causal determinant of persistent neural learning. By deliberately altering compatibility between matched neural states, the authors show that persistent learning can be measured under a common retention requirement across different learning directions, architectures, and data modalities. The study reveals that learning rules vary in how efficiently they exploit compatibility, and that retention constraints and finite updates limit what can be stored, with nonlinear geometry ultimately preventing full compatibility realization.
By Hossein Javidnia
The paper introduces Generalized Residual Closure (GRC), a framework that models learning as a recursive closure of future‑relevant discrepancies, enabling a learner to acquire new capabilities while preserving prior responsibilities. GRC identifies two primitive internal learning modes—transformation within a representation and revision of the representation itself—and provides a local tangent decomposition, a criterion for when representation revision is needed, and a necessary‑and‑sufficient condition for stability‑plasticity compatibility in an affine model. The authors further demonstrate how GRC supports safe, constrained updates that reduce residuals, extend representation adequacy to future updates, and connect the framework to optimization, machine learning, and control theory.
whyItMatters":"The framework offers a principled way to balance stability and plasticity, ensuring continued learning without regression of previously acquired capabilities."
By Dongxu Li, Yinuo Zhang, Hongyu Zhang, Feng Tian
arXiv:2606. 03532v1 Announce Type: cross Abstract: Self on-policy distillation trains a student policy against a teacher derived from its own parameter history, yet the teacher's update schedule -- which governs the \emph{temporal coupling} between teacher and student -- has not been systematically studied as a stability variable.
By Haowei Guo, Baolong Bi, Ruicheng Zhang, Bingqian Sun, Wentao Zhang
Steering vectors (SVs) are residual-stream directions that can shift LLM behavior when added to hidden states during inference. This study investigates whether the same SVs can guide parameter updates during post‑training, exploring supervised learning (SFT and distillation) and reinforcement learning with verifiable rewards (RLVR). The authors find that SV injection improves performance over plain SFT and distillation, achieving up to a 12% gain in certain distillation settings, and that applying a correctness‑contrast direction as advantage repair in RL recovers learning signals for held‑out math benchmarks. They also identify failure boundaries related to code generation, cross‑family layer transfer, and inference‑time application on SV‑trained checkpoints.
By Tiejin Chen, Maunil Vyas, Huaiyuan Yao, Hua Wei
arXiv:2609.38768v1 Announce Type: cross
Abstract: Prior work has shown that neural networks exhibit implicit biases toward low-complexity structure (e.g., spectral bias), memorization dynamics, and c...
By Fujie Gao, Zuyue Zhang, Gang Sun