Multimodal code intelligence connects the symbolic world of programs with the visual world their execution creates. Interest in this capability is burgeoning, as models are increasingly asked to produce charts, webpages, animations, and semi-structured documents. Recently, proprietary models have pushed far ahead, prod...
Qiushi Sun, Jingyang Gong, Yang Liu et al.· 0 citations
A multimodal agent must decide whether visual evidence is worth acquiring, which candidate or resolution to query, and when further interaction can no longer change its decision. We study this problem under a strict free-metadata/paid-observation boundary, stochastic vector costs, and a frozen execution trace, and trea...
Miao-Bo Hu, Shu-Hao Hu, Xiao-Bo Guo et al.· 0 citations
On-policy self-distillation provides dense supervision for reasoning models via a privileged self-teacher. However, this supervision reflects local token preferences rather than the expected return of an unguided student continuation. Recent methods combine this signal with GRPO. They use GRPO for update directions and...
Fei Wu, Zhen-Rong Zhang, Jian-Shu Zhang et al.· 0 citations
Generative super-resolution (SR) can recover sharp detail in computed tomography (CT) images from severely downsampled observations, but the missing high-frequency content is not uniquely determined by the low-resolution input. Consequently, visually plausible reconstructions may contain anatomically unsupported vascul...
Hai-Rong Ni, Jiahe Huang, Ming-Hao Guo et al.· 0 citations
Reach audiences
Advertise in front of researchers, engineers, and readers.
Opponent histories can mislead after behavioral shifts. SafeAdapt-Poker formulates budgeted adaptation under opponent shift as a dual-counterfactual sequential-control problem using only information available before each decision. At each decision the controller asks two distinct questions: what local downside a deviat...
Miao-Bo Hu, Shu-Hao Hu, Xiao-Bo Guo et al.· 0 citations
Context can improve large language model performance, but these gains may be lost after context removal. Inspired by biological systems consolidation, we formulate Memory Consolidation (MC) as internalizing these gains while preserving non-target capabilities, connecting context-specified editing, skill learning, and b...
Jintian Zhang, Feng He, Xinnian Liang et al.· 0 citations
LSPO is introduced, which establishes this finite, cost-adjusted transition as the primary unit of learning, and each textual answer is augmented with a learned continuous coordinate, forming a lifted state that an energy function evaluates and orders.
Large language models increasingly generate multiple retrieval actions jointly, while downstream systems often provide only one reward for the resulting set. Broadcasting this reward provides a valid score-function gradient estimator for the expected set-level objective, but does not determine a coalition-dependent wit...
Xinyi Cheng, Tianfu Zhang, Jiwen Li et al.· 0 citations
In-context control of a model's own activations might not yet be a threat to activation monitors, but it exists in a few models and needs to be tracked as models evolve.
Davide Baldelli, Robyn Kozierok, Mirko Bronzi et al.· 0 citations
AtomSearch is introduced, which keeps revision actions local while using subsequent skill states to guide search, and shows how local rule edits can support the exploration of useful revision paths and yield compact skills, with additional offline evaluations used to explore alternative states.
Jiang-Shuai-Chen-Rui-,-Zhang-Rong-,-Wang-Gui-Fang- Tong, Sen Deng, Ming-Pei Peng et al.· 0 citations
Adapting a single vision-language model to medical image parsing requires training choices that reflect the demands of each task. We study task-specific adaptation with Qwen3.5-9B for FLARE 2026 Task 3, covering six task families and seven imaging modalities. Our system couples a shared, frozen 4-bit backbone with LoRA...
Taha Sabir, Alberto Montero Molina, Thinh Quang Ly et al.· 0 citations
Training LLM agents on long-horizon interactive tasks with reinforcement learning typically relies on human-annotated tasks and reliable verification, both expensive to construct for stateful environments. Self-play offers an alternative, but must generate achievable, compositional tasks and provide feedback beyond bin...
Tianyu Liu, Davide Belli, Bence Major et al.· 0 citations
