Papers arxiv:2605.28819 PEFT-Arena: Understanding Parameter-Efficient Finetuning from a Stability-Plasticity Perspective Published on May 27 Submitted by Yangyi, Huang on May 28 CUHK SphereLab Authors: Yangyi Huang , , , , , , Weiyang Liu Abstract Parameter-efficient fine-tuning methods exhibit var… …
… However, these algorithms suffer from an exploration-stability dilemma . Pure IS often leads to catastrophic training instability, while standard clipping mechanisms used to mitigate this instability strictly constrain the policy update budget . …
… We disentangle these effects through a controlled decomposition, and ask which is responsible for semantic control and which governs downstream stability. …
… Enforcing Eikonal regularization ensures non-vanishing gradients near the collision boundary, improving numerical stability and robustness of the optimization process. …
… We empirically and theoretically show that effective test-time evolution requires jointly adapting both axes, but on different time scales: capabilities should update rapidly to handle emerging subtasks, while the topology should evolve more slowly to preserve coordination stability. …
… Evaluated on public benchmarks and our in-house RealMobile, Xiaomi-GUI-0 achieves 72.0% success on RealMobile and 78.9% on AndroidWorld, while substantially improving execution stability and abnormal-state recognition in real-world tasks. …
… This is the operationalization of "PT-breaking ↔ causal order." iii η-norm drift. …
…EgoForce operates across fisheye, perspective, and distorted wide-FOV camera model s using a single unified network. Our approach combines a differentiable forearm representation that stabilizes hand pose, a unified arm-hand…
…Ziyu Guo , , , Abstract ATLAS presents a visual reasoning framework that combines agentic operations and latent representations using functional tokens, enabling efficient training and improved performance on complex benchmarks. AI-generated summary Visual…
… It operationalizes this latent-refinement architecture through four core components: 1 block decomposition , guided by staged representation dynamics ; 2 an input-dependent selective gate to mitigate hidden-state drift ; 3 random deep supervision for memory-efficient learning over long recursive ho… …