Not All Synthetic Data Is Yours to Learn From

Can a language model improve from plain text sampled from itself, with no prompts, no teacher, no verifier, and no reward model? Yes, but only when the synthetic corpus is compatible with the student, a relational property of the source-student pair rather than an intrinsic property of the data. We call this the latent capability resurfacing hypothesis: weak self-training can amplify capabilities already present in the pretrained model, but only under this compatibility condition. We study this in the minimal setting of prompt-free unconditional self-training, where base language models are fine-tuned on text generated from the BOS token alone, with no task specification or external supervision. We report three findings. First, synthetic utility is relational rather than intrinsic: self-generated data is the most effective source, same-lineage transfer outperforms stronger but differently trained sources, and cross-family transfer is substantially weaker. Second, common intrinsic proxies fail: neither benchmark-level semantic similarity nor average per-token likelihood under the student predicts which corpora help. Third, this regime produces a surprising byproduct. In controlled Pythia experiments, capability and verbatim memorization decouple: benchmark utility is preserved or improved while held-out exact-match extraction drops by over 95 percent, with no forget set, privacy objective, or targeted unlearning. Together, these results suggest that prompt-free self-training works by amplifying what the student already knows, not by importing structure from the data. They also reveal a regime in which capability and verbatim memorization can be separated without any explicit unlearning objective.

Deconstructing Spatial Complexity: Hierarchical Decomposition for LLM Spatial Reasoning

LLMs have shown remarkable proficiency in general language understanding and reasoning. However, they consistently underperform in spatial reasoning that severely limits their application, particularly in embodied intelligence. Inspired by the success of hierarchical reinforcement learning, this paper introduces a novel method for hierarchical task decomposition in LLM spatial reasoning. Our approach guides LLMs to decompose complex tasks into manageable sub-tasks by identifying key intermediate states and generating simplified sub-environments. However, we identify that LLMs often fail to derive optimal intermediate states due to their insufficient spatial prior, leading to sub-optimal task decomposition. To address this limitation and enhance its planning capability, we propose the MCTS-Guided Group Relative Policy Optimization (M-GRPO), where we reformulate the UCT formula by incorporating the LLM's prior predictive probabilities alongside its epistemic uncertainty. Furthermore, we implement a more fine-grained advantage function, enabling the model to learn optimal path planning. Experimental results demonstrate that our method substantially improves LLM performance on spatial tasks, including navigation, planning, and strategic games, achieving state-of-the-art results. This work paves the way for LLMs in real-world applications.

Capture-Calibrate-Coach: A Graph-Based Framework for Knowledge Monitoring Estimation and Adaptive Feedback

Effective learning support requires understanding not only what learners know but also how accurately they perceive their own understanding. This metacognitive dimension, known as knowledge monitoring, fundamentally influences self-regulated learning, yet this dimension remains underexplored in current systems. This paper introduces the Capture-Calibrate-Coach (3C) framework for adaptive learning support. The Capture phase extracts learners' perceived knowledge states from open-ended self-reports to construct a heterogeneous graph linking learners and knowledge concepts. The Calibrate phase applies a heterogeneous graph neural network to infer latent perceived states for concepts not explicitly mentioned, enabling systematic knowledge monitoring assessment. The Coach phase classifies learners into five metacognitive patterns and delivers personalized feedback addressing both knowledge gaps and calibration errors. Evaluation with 684 students demonstrates 85.21% AUC in predicting latent perceived states, significantly outperforming baseline methods. A user study with 47 participants shows positive reception of feedback quality, with participants particularly valuing concrete feedback on knowledge gaps and actionable study guidance. These findings advance AI-based learning support toward metacognitive teammates that foster accurate self-awareness while supporting knowledge growth.

InsightTok: Improving Text and Face Fidelity in Discrete Tokenization for Autoregressive Image Generation

Text and faces are among the most perceptually salient and practically important patterns in visual generation, yet they remain challenging for autoregressive generators built on discrete tokenization. A central bottleneck is the tokenizer: aggressive downsampling and quantization often discard the fine-grained structures needed to preserve readable glyphs and distinctive facial features. We attribute this gap to standard discrete-tokenizer objectives being weakly aligned with text legibility and facial fidelity, as these objectives typically optimize generic reconstruction while compressing diverse content uniformly. To address this, we propose InsightTok, a simple yet effective discrete visual tokenization framework that enhances text and face fidelity through localized, content-aware perceptual losses. With a compact 16k codebook and a 16x downsampling rate, InsightTok significantly outperforms prior tokenizers in text and face reconstruction without compromising general reconstruction quality. These gains consistently transfer to autoregressive image generation in InsightAR, producing images with clearer text and more faithful facial details. Overall, our results highlight the potential of specialized supervision in tokenizer training for advancing discrete image generation.

RecGPT-Mobile: On-Device Large Language Models for User Intent Understanding in Taobao Feed Recommendation

Predicting a user's next search query from recent interaction behaviors is a critical problem in modern e-commerce systems, particularly in scenarios where user intent evolves rapidly. Large Language Models (LLMs) offer strong semantic reasoning capabilities and have recently been adopted to enhance training data construction for next-query prediction. However, due to resource constraints on mobile devices, existing applications are deployed on cloud servers, resulting in high inference costs. In this paper, we propose RecGPT-Mobile, a framework that designs a lightweight LLM-based intent understanding agent to improve recommendation quality in mobile e-commerce scenarios. By deploying LLMs directly on mobile devices, our approach can capture evolving interests of users more quickly and adjust the recommendation results in real time. Extensive offline analyses and online experiments demonstrate that our method significantly improves the accuracy of recommendation results, laying a practical path for LLM deployment in production-scale recommendation systems on mobile devices, as well as a scalable solution for integrating LLMs into real-world next-query prediction systems.

Causality-Encoded Diffusion Models for Interventional Sampling and Edge Inference

Standard diffusion models are flexible estimators of complex distributions, but they do not encode causal structures and therefore do not by themselves support causal analysis. We propose a causality-encoded diffusion framework that incorporates a known directed acyclic graph by training conditional diffusion models consistent with the graph factorisation. The resulting sampler approximately recovers the observational distribution and enables interventional sampling by fixing intervened variables while propagating effects through the graph during reverse diffusion. Building on this interventional simulator, we develop a resampling-based test for directed edges that generates null replicates under a candidate graph. We establish convergence guarantees for observational and interventional distribution estimation, with rates governed by the maximum local dimension rather than the ambient dimension, and prove asymptotic control of type I error for the edge test. Simulations show improved interventional distribution recovery relative to baselines, with near-nominal size and favourable power in inference. An application to flow cytometry data demonstrates practical utility of the proposed method in assessing disputed signalling linkages.

SAGER: Self-Evolving User Policy Skills for Recommendation Agent

Large language model (LLM) based recommendation agents personalize what they know through evolving per-user semantic memory, yet how they reason remains a universal, static system prompt shared identically across all users. This asymmetry is a fundamental bottleneck: when a recommendation fails, the agent updates its memory of user preferences but never interrogates the decision logic that produced the failure, leaving its reasoning process structurally unchanged regardless of how many mistakes it accumulates. To address this bottleneck, we propose SAGER (Self-Evolving Agent for Personalized Recommendation), the first recommendation agent framework in which each user is equipped with a dedicated policy skill, a structured natural-language document encoding personalized decision principles that evolves continuously through interaction. SAGER introduces a two-representation skill architecture that decouples a rich evolution substrate from a minimal inference-time injection, an incremental contrastive chain-of-thought engine that diagnoses reasoning flaws by contrasting accepted against unchosen items while preserving accumulated priors, and skill-augmented listwise reasoning that creates fine-grained decision boundaries where the evolved skill provides genuine discriminative value. Experiments on four public benchmarks demonstrate that SAGER achieves state-of-the-art performance, with gains orthogonal to memory accumulation, confirming that personalizing the reasoning process itself is a qualitatively distinct source of recommendation improvement.

Rethinking AI Hardware: A Three-Layer Cognitive Architecture for Autonomous Agents

The next generation of autonomous AI systems will be constrained not only by model capability, but by how intelligence is structured across heterogeneous hardware. Current paradigms -- cloud-centric AI, on-device inference, and edge-cloud pipelines -- treat planning, reasoning, and execution as a monolithic process, leading to unnecessary latency, energy consumption, and fragmented behavioral continuity. We introduce the Tri-Spirit Architecture, a three-layer cognitive framework that decomposes intelligence into planning (Super Layer), reasoning (Agent Layer), and execution (Reflex Layer), each mapped to distinct compute substrates and coordinated via an asynchronous message bus. We formalize the system with a parameterized routing policy, a habit-compilation mechanism that promotes repeated reasoning paths into zero-inference execution policies, a convergent memory model, and explicit safety constraints. We evaluate the architecture in a reproducible simulation of 2000 synthetic tasks against cloud-centric and edge-only baselines. Tri-Spirit reduces mean task latency by 75.6 percent and energy consumption by 71.1 percent, while decreasing LLM invocations by 30 percent and enabling 77.6 percent offline task completion. These results suggest that cognitive decomposition, rather than model scaling alone, is a primary driver of system-level efficiency in AI hardware.

An Adaptive Antenna Impedance Matching Method via Deep Reinforcement Learning

Adaptive impedance matching between antennas and radio frequency front-end modules is critical for maximizing power transmission efficiency in mobile communication systems. Conventional numerical and analytical methods struggle with a trade-off between accuracy and efficiency, while deep neural network (DNN)-based supervised learning approaches rely heavily on large labeled datasets and lack flexibility for dynamic environments. To address these limitations, this paper proposes a deep reinforcement learning (DRL)-based approach for adaptive impedance matching. First, we model the impedance tuning problem as an optimal control problem, proving the feasibility of solving the optimal control law via reinforcement learning. Then, we design a tailored DRL framework for impedance tuning, which employs a compact state representation that integrates key frequency characteristics and matching quality metrics. Additionally, this framework incorporates a piecewise reward function that accounts for both matching accuracy and tuning speed. Furthermore, a test-phase exploration mechanism is introduced to enhance tuning stability, which effectively reduces local optimal trapping and high-frequency tuning variance. Experimental results demonstrate that the proposed method achieves superior performance in terms of tuning accuracy, efficiency, and stability compared with conventional heuristic and gradient-based methods, making it promising for practical impedance tuning systems.

TAMEn: Tactile-Aware Manipulation Engine for Closed-Loop Data Collection in Contact-Rich Tasks

Handheld paradigms offer an efficient and intuitive way for collecting large-scale demonstration of robot manipulation. However, achieving contact-rich bimanual manipulation through these methods remains a pivotal challenge, which is substantially hindered by hardware adaptability and data efficacy. Prior hardware designs remain gripper-specific and often face a trade-off between tracking precision and portability. Furthermore, the lack of online feasibility checking during demonstration leads to poor replayability. More importantly, existing handheld setups struggle to collect interactive recovery data during robot execution, lacking the authentic tactile information necessary for robust policy refinement. To bridge these gaps, we present TAMEn, a tactile-aware manipulation engine for closed-loop data collection in contact-rich tasks. Our system features a cross-morphology wearable interface that enables rapid adaptation across heterogeneous grippers. To balance data quality and environmental diversity, we implement a dual-modal acquisition pipeline: a precision mode leveraging motion capture for high-fidelity demonstrations, and a portable mode utilizing VR-based tracking for in-the-wild acquisition and tactile-visualized recovery teleoperation. Building on this hardware, we unify large-scale tactile pretraining, task-specific bimanual demonstrations, and human-in-the-loop recovery data into a pyramid-structured data regime, enabling closed-loop policy refinement. Experiments show that our feasibility-aware pipeline significantly improves demonstration replayability, and that the proposed visuo-tactile learning framework increases task success rates from 34% to 75% across diverse bimanual manipulation tasks. We further open-source the hardware and dataset to facilitate reproducibility and support research in visuo-tactile manipulation.