Skills let language agents reuse experience across tasks as natural-language documents. Recent work evolves skills from interaction, but existing methods have limitations when skills evolve online. Prior methods update a skill offline from a batch of trajectories collected over many tasks. Some then adopt the new skill without any check, which can degrade the skill over time. Others keep it only if it improves performance on a held-out validation set, which is too slow and costly to run after every update. We introduce EvoGuard, a pairwise skill selector that mitigates skill degradation at low latency and cost. We first run the agent iteratively to build pools of skills and score each skill on a private test set to obtain utility labels. We pair the skills and use rejection sampling to keep only pairs with a clear utility gap. The selector is trained on the logits of two label tokens to pick the better skill. At deployment it decides in a single forward pass, which is significantly faster than evaluation on the private test set. Compared with adopting every new skill, EvoGuard improves success by 11.0% on ALFWorld, 5.9% on ScienceWorld, and 2.7% on LiveCodeBench.
Multivariate time series anomaly detection (MTSAD) relies on modeling both temporal dynamics and inter-sensor dependencies to identify complex system faults. Although graph-based detectors explicitly model sensor topologies, they train task-specific networks from scratch on limited target-domain data, remaining vulnerable to spurious correlations and unable to reuse transferable priors. While pretrained Graph Foundation Models (GFMs) offer transferable topological priors and anomaly discriminability, directly adapting them to MTSAD is hindered by cross-modal feature-topology mismatch, unreliable pseudo-anomaly supervision, and decision distortion during node-to-window aggregation. To fill these gaps, we propose a novel adaptation framework, Gfm4Mtsad, which unleashes pretrained GFMs for MTSAD. We introduce the Structure-and-Information Contract Alignment (SICA) module to bridge continuous time series and the pretrained GFM space. This module harmonizes principal subspace token projection, normal-prior-guided graph construction, and residual recovery to suppress spurious edges while retaining fine-grained temporal details. We propose a Mechanism-Aligned Pseudo-Anomaly Screening strategy (MAPS) to steer unsupervised learning. It applies localized dependency breaks while preserving empirical residuals, selecting hard negatives near the normal boundary for targeted supervision. We develop a Tail-Gated Evidence Calibration mechanism (TGEC) to eliminate node-to-window aggregation distortion. It calibrates multi-granular graph responses against normal operational references, strictly preserving benign score rankings while selectively refining extreme anomaly candidates in the tail distribution for reliable window detection and interpretable sensor evidence. Extensive evaluations with two GFM backbones demonstrate that Gfm4Mtsad achieves superior detection performance over existing baselines.
Chain-of-Thought (CoT) reasoning substantially improves Large Language Models (LLMs), yet its out-of-distribution (OOD) generalization mechanism remains theoretically underexplored. We study CoT as compositional OOD generalization: the target task lacks complete reasoning trajectories or direct input-output pairs, while training provides only in-distribution (ID) atomic step data. We show that CoT generalizes by repeatedly applying a shared single-step predictor to compose reusable ID transitions. Theoretically, let $T$ denote the number of reasoning steps in the target task, $n$ the number of training samples per sub-task, and $k$ the number of output tokens per step; then the OOD task risk is controlled by the sum of the ID sub-task risks, scaled by a coefficient of order $\mathcal{O}(k)$ that is independent of $T$, although the summed risk itself may still grow with $T$. A Rademacher-complexity analysis further shows that the estimation term is of order $\mathcal{O}(\sqrt{T/n})$, rather than linear in $T$, due to predictor sharing under a uniform capacity control on the shared class. For autoregressive Next-Token Prediction, the final-task risk is bounded by the best-in-class step-level risk scaled by the same coefficient of order $\mathcal{O}(k)$, plus an estimation term of order $\widetilde{\mathcal{O}}\big(k\sqrt{T/n}\big)$. A parity-coupled construction further shows that, for the sub-task empirical risk minimizer, the expected final-task risk is at least the best in-class risk plus $\Omega(\sqrt{T/n})$ when $T\le n$. Experiments on Atom-Task and GSM8K support the theory: CoT enables short-to-long and compositional generalization, while explicit state tracking is crucial. Our code is available at https://github.com/maxliang114514/CoT-OOD.
We show that a hardwired distributed processing unit (HDPU) can take a neural decision at the sensor and so remove the inference from the host. Built for zebra-crossing detection on a low-cost Artix-7 FPGA, the unit decides on every camera frame in a fixed 2,269,477 clock cycles (34.04 ms depends upon Fpga clk ), inside the 37.49-ms frame period, and on the board it returns the same scores and decisions as its bit-exact software model on all 639 test images. The whole detection path, from camera pixels to decision, takes 1,814 LUTs, with no processor, no operating system and no full-resolution frame store. When the unit answers instead of reporting "not sure", 95.7% of its answers on street photographs are correct, and a hardware contrast stage lifts accuracy on camera-like dashcam frames from 47.7% to 78.2%. The host is spared all of the inference: 8.13 ms of a core in every frame period, where a full-precision ResNet-18 would need more than a frame period even on 16 threads. It also receives 16 bytes per frame instead of 614,400, 38,400 times less. The unit is an 80x60-pixel pipeline feeding a four-layer binarized CNN with 93,408 one-bit weights, verified bit-exactly from a Python model through RTL simulation to the board. Earlier HDPUs did pre-processing (edges, corners, steering); this one shows that the decision itself can move to the sensor.
Reliable channel estimation and symbol detection in 5G New Radio (NR), as well as emerging 6G radio systems, rely on demodulation reference signals (DMRS). The density and placement of these pilots on the orthogonal frequency-division multiplexing (OFDM) time--frequency resource grid affect both receiver reliability and spectral efficiency. Sparse DMRS layouts increase the time--frequency distance between pilot observations and data resource elements (REs), which can make pilot information harder to exploit with purely local processing. This paper proposes \textbf{PARGAR (Pilot-Aware Graph Attention Receiver)}, a graph-attention neural receiver that represents the OFDM resource grid as a graph for bit detection. The graph combines local time--frequency edges with pilot-anchor links from data REs to selected pilot REs, and we further study a channel-scaled pilot-anchor rule based on approximate coherence time and coherence bandwidth. Using Sionna-generated OFDM simulations, we evaluate \textsc{PARGAR} against DeepRx, a convolutional neural network (CNN)-based receiver, across dense, two-symbol sparse, and single-symbol sparse DMRS layouts under channel-mismatch scenarios. Under the considered SISO uncoded setup, channel-scaled \textsc{PARGAR} explicitly injects pilot-aware inductive bias and yields lower BER in selected sparse-pilot scenarios while remaining competitive under dense pilots.
Pansharpening has been extensively studied as a fundamental image fusion task in remote sensing. However, most existing methods focus on improving fusion quality under the implicit assumption that high-resolution observations have already been fully downlinked to the ground. As a result, existing models are highly vulnerable to perturbations introduced by lossy compression used for downlink transmission. To address this limitation, we present, to the best of our knowledge, the first transmission oriented pansharpening (TOP) framework, which unifies compact latent coding, onboard pansharpening, and efficient downlink transmission within a single satellite-toground pipeline. TOP encodes remote-sensing observations into a compact binary latent space, where a lightweight onboard fusion module can efficiently predict the fused binary latent code. Since only the compact fused code is downlinked, the transmission payload is substantially reduced. Overall, TOP enables low-payload downlink and lightweight onboard processing while maintaining high fusion quality under severe bandwidth constraints. Experiments across multispectral and hyperspectral settings show that TOP consistently achieves superior fusion quality at higher compression ratios than representative pansharpening methods paired with standard lossy codecs. For example, on GaoFen-2, TOP attains a fusion PSNR of 31.4 dB at a downlink compression ratio of 50x. This design also points toward transmissionoriented processing for other remote-sensing tasks. The code will be made available at https://github.com/JingZe-Liang/TOP.git.
Partially interpretable models extend simple interpretable models with the option to defer complicated inputs to an accurate but uninterpretable black-box model. By deferring more inputs, the model gets more accurate; by deferring fewer, it gets more interpretable. Existing procedures to train such partially interpretable models build on linear models or decision rules. We propose three algorithms for decision trees: a partially interpretable extension of the existing Branches algorithm, which performs exact search to optimize the tree; a conversion algorithm that we run on top of standard CART, which is computationally more efficient; and a generalized CART whose splitting criterion takes into account the option to defer while growing the tree. Perhaps surprisingly, the performance of the three algorithms is similar across many realworld datasets. We can explain this through a formal bound on the potential gain of smart tree construction, provided the black box accuracy is sufficiently homogeneous across the domain. We also construct a synthetic dataset where the accuracy of the black box is deliberately inhomogeneous. There, our deferral-aware generalization of CART does outperform regular CART and matches the performance of exact search. Finally, we evaluate an application to concept bottleneck models. Partially interpretable decision trees can replace uninterpretable neural networks after the bottleneck. On undeferred inputs, the model is fully interpretable, and deferring preserves accuracy.
This work presents UGD-FM, a novel unaligned image guided denoising framework based on cross-modal conditional flow matching. Unlike previous approaches that assume spatially aligned inputs or only handle small-baseline rectified pairs, we address a more challenging setting where severe target noise and large-range two-dimensional cross-modal misalignment coexist. Specifically, we formulate guided denoising as a progressive flow matching process, allowing image restoration and aligned guidance aggregation to mutually reinforce each other. To support efficient few-step inference, we introduce a stage-focused and inference-consistent training strategy that focuses on task-critical denoising stages, and mitigates the training-inference mismatch. We further propose adaptive sparse guidance propagation for efficient and reliable large-range matching, regularized by perceptual and geometric auxiliary losses. Experiments demonstrate that UGD-FM achieves state-of-the-art denoising performance with only one or two inference steps, and is compatible with modern single-image restoration backbones.
Multimodal large language models (MLLMs) have demonstrated remarkable progress on cross-modal reasoning tasks, yet they continue to suffer from hallucinations in long-chain reasoning processes. In this work, we revisit hallucination in MLLMs from the perspective of reasoning generation dynamics. We argue that entropy is not merely a static measure of uncertainty but a structured signal that reflects how modality information is integrated throughout decoding. Through systematic analysis, we identify two characteristic entropy patterns associated with hallucination behaviors: (1) a generally high-confidence regime corresponds to premature convergence to inappropriate multimodal interactions; and (2) local abrupt entropy transitions are associated with reasoning instability and hallucination onset. Building on these observations, we propose an entropy-aware self-introspective framework for MLLM reasoning. Our method continuously tracks entropy dynamics during decoding and performs targeted visual re-alignment when suspicious hallucination states are detected, encouraging the model to re-examine salient visual evidence at critical reasoning stages. Extensive experiments across six multimodal reasoning benchmarks spanning nine distinct evaluation settings demonstrate that our approach delivers robust improvements in reasoning accuracy and hallucination robustness, while providing new insights into the generation dynamics underlying MLLM failures. The code is available at: https://github.com/jden4524/SelfRetrospectiveDecoding.
Long-term problem-solving processes, such as scientific research and software development, proceed along an evolving trial-and-error trajectory toward a central goal. We term these \textit{exploration tasks}, which require models to remember not only facts but also how solutions are tested against intermediate outcomes and early attempts are continually revised or revived. However, existing long-term memory research focuses on factual recall and updates, overlooking the interconnected nature of exploration. We observe that exploration processes naturally form graphs, with branching approaches linked by revisions and reconsiderations. Motivated by this, we propose \textsc{LExMem}, a graph-guided benchmark designed to evaluate \underline{\textsc{Mem}}ory in \underline{\textsc{L}}ong-term \underline{\textsc{Ex}}ploration tasks. We model exploration processes as graphs and introduce an evolutionary mechanism to increase their structural complexity, yielding 224 dialogues and 2,018 question-answer instances derived from graph structures. Building on the same insight, we propose \textsc{LExGraph}, a graph-based memory framework that incrementally organizes dialogues into interconnected memory nodes and retrieves evidence through hybrid retrieval and neighbor expansion. Extensive experiments demonstrate that \textsc{LExMem} poses a significant challenge to existing retrieval- and memory-based systems, whereas \textsc{LExGraph} achieves superior performance and generalizes effectively to external benchmarks. Codes and datasets are available at https://github.com/mbzuai-nlp/LExMem.
The protection of mature, multi-stage Face Recognition Systems is crucial to system providers. This technical report targets the threat model most relevant to industry practitioners: although model training is performed in-house, training data is sourced from possibly-compromised third parties. Given this setting, we present three practical defense "tricks" against data poisoning-based backdoor attacks. The first operates at training-time, detecting poisoned identities during training or prior to a pipeline's deployment. The last two operate at inference-time, filtering suspicious inputs before an answer can be sent back to a potential attacker. Each method is presented with their motivation, algorithms and computational impact, and evaluated on realistic face recognition workloads. Our goal is to provide actionable hardening techniques to system providers.
We study zero-shot coordination (ZSC), where independently developed agents must coordinate at test time. While ZSC has been well studied in the RL literature, far less is known about the performance of LLM agents despite their increasing deployment in such settings. Existing work on LLM coordination often relies on specialised scaffolding that independently developed agents are unlikely to share in practice. Moreover, evaluations on complex environments (e.g., Hanabi) make it difficult to pinpoint the sources of coordination failure. In contrast, we focus on simple, general-purpose scaffolds in minimal environments designed to isolate specific coordination challenges. Our results show that even in these controlled settings, frontier LLM agents struggle to coordinate, largely due to limited understanding of the coordination problem and weak reasoning about their partner's beliefs. While LLMs introduce semantic information as an additional axis for coordination, they nonetheless fail to exploit this structure effectively. Towards this, we propose Coordination-friendly definitions (CFDs) as a principled approach for enabling robust coordination among LLM agents. Finally, we show that CFDs can be discovered automatically, removing the need for manual engineering. Code is available at https://github.com/ahayler/zsc-llm-agents.
Achieving both high pronunciation accuracy and fine-grained controllability remains a central challenge in Japanese text-to-speech (TTS), as Japanese is an unsegmented language mixing logographic Kanji and phonographic Kana with extensive polyphony. Traditional cascaded systems convert text into phoneme sequences via a front-end, ensuring pronunciation control but discarding lexical semantics and prosodic cues in surface orthography. Conversely, recent zero-shot TTS models directly consume raw orthographic text to improve naturalness, yet suffer from polyphone errors and lack explicit pronunciation control. To bridge this gap, we propose a two-stage framework for flow-matching TTS. First, we investigate Ruby text representations with orthography perturbation to jointly preserve surface forms and target pronunciations for reliable control. Second, we introduce On-Policy Distillation (OPD) on large-scale text-only corpora to enhance pronunciation accuracy on raw orthographic text without paired speech. On the Joyo-Kanji-Yomi benchmark, our model achieves 97.13% Accuracy, 2.02% Target PER, 0.70% Sentence PER, and 4.21% Sentence CER under raw text while retaining strong controllability.
QA-based text embeddings represent texts as vectors consisting of 1/0 answers to yes/no questions. Although these embeddings provide detailed information for interpretation and achieve high task performance, they often require humans to examine a large number of questions. Furthermore, the relationship between the number of questions and the appropriate level of question granularity has not been sufficiently studied. To address these challenges, we propose a method for constructing hierarchical questions for text embeddings. As the hierarchy deepens, questions become more specific and their number increases. During interpretation, we reduce the number of questions that humans need to examine by replacing questions at deeper levels with aligned questions at shallower levels. The question hierarchy is constructed through recursive clustering, contrastive prompting, and question alignment. Experimental results on general-purpose tasks show that our method improves interpretability while maintaining task performance. It is also found that, when only a limited number of questions are available, questions at shallower levels tend to achieve higher task performance than those at deeper levels, even with the same number of questions.
Non-prefix KV-cache reuse encodes retrieved chunks independently and concatenates their caches. To find where a small repair budget should be spent, we vary the chunks around the answer-bearing one (the target) across three 7B-8B models and four datasets of dialogue memory, open-domain QA and API documentation. Reuse loses 19-100% of the context-dependent questions that full prefill answers, and it loses more when the other chunks are of the same kind as the target. Because the recoverable damage is spread over these other chunks, restoring a few of them recovers little, and the budget is better spent on the target. This works through contextual compatibility: recomputing the target against the reused cache beats inserting its full-prefill KV in all seven settings tested, although the recomputed KV stays far from the full-prefill KV. Question attention over reused caches, however, selects the target in at most 18.8% of development questions. A retriever finds it more often, and recomputing its top chunk after moving it next to the question significantly beats InfoFlow-style reordering at the same 5% budget in six of nine open-domain QA and API-documentation settings. On dialogue memory, though, recomputing chunks never significantly beats the token-level selector chosen for each setting.
Typed decision models offer structured choices, scores, and probabilities for low-cost automated evaluation. We present a controlled audit of presentation sensitivity in Jev, Clef Flash, and Perplexity Decider across native choice and score interfaces. Across six datasets with 9,000 multiply annotated humor, helpfulness, and summarization items, we collect 702,000 judgments under interventions that vary positions, identifiers, authorship claims, endorsements, decision framing, and rating rubrics while preserving the evaluated texts. Jev's second-position probability advantage reaches 6.11 percentage points on summaries; reversing reader endorsements changes Perplexity's selected humor text on 40.40% of pairs; and rejection framing reduces Clef's summary agreement with human preferences by 23.36 points. Rating formats also shift score levels while often preserving rankings. Across these findings, repeatability, presentation robustness, and human alignment emerge as distinct properties of an evaluator. Reliable use therefore requires validating the decisions and score interpretations on which an evaluation pipeline depends.
Dooge asked whether hydrology has laws at the catchment scale. We ask whether a low-dimensional system built under physical laws can match the accuracy of neural networks that outperform process-based models on the same inputs. Three laws constrain a lumped catchment model. Conservation makes it a flux-connected storage network, dissipation fixes the sign of every flux and aggregation determines how the unresolved composition enters the fluxes, through effective laws over threshold distributions and a memory of the forcing. These laws admit four kinds of refinement. A parsimonious model under them, CaRT (11 parameters per catchment), is refined in a loop assisted by AI agents. Each refinement addresses an observed deficit through a hypothesis from the literature. The result, CaRT+, has fourteen states, 35 parameters per catchment and no daily neural network. Most refinements follow from aggregation over thresholds or from variables that carry no water but modulate fluxes. On 602 British and 593 US gauged catchments, single CaRT+ models reach median Nash-Sutcliffe efficiencies of 0.863 and 0.765 in held-out years. A stateful LSTM reaches 0.869 and 0.751 and an entity-aware LSTM 0.862 and 0.754, while three-seed LSTM ensembles remain ahead of the CaRT+ ensemble by 0.003-0.012 in median paired NSE. On 505 drier Australian gauged catchments unused in development, single CaRT+ models reach 0.778 against 0.762 and 0.753 for the two LSTMs. The results suggest that catchment-scale laws exist as laws of form, with the constitutive families shared and only their parameters learned.
Loop closure detection conventionally cascades separate models for global retrieval, geometric verification, and relative pose estimation, leading to redundant computation and disconnected visual representations. We propose UniLoop, a unified framework that reuses the hierarchical representations of the pretrained geometric foundation model throughout this pipeline. A shared image encoder produces features for compact global descriptors; the same features are passed to the cross-view decoder for pairwise verification; and the original geometric predictions provide relative camera constraints for pose-graph optimization. Experiments on place recognition, loop verification, and relative pose estimation show that UniLoop improves loop-closure robustness and accuracy while reducing redundant backbone computation relative to modular pipelines. UniLoop achieves the best Recall@10 on MSLS (96.6) and Pitts30k (99.3) among the compared methods, while its verification branch reaches 95.63 AP and 96.29 ROC-AUC, surpassing the existing geometric verification baselines.
Autoregressive video models rely on historical key-value (KV) caches to preserve scene consistency over extended rollouts, making KV storage and repeated attention over history a growing memory and computational bottleneck. Aggressive compression can degrade generation quality, while reconstructing full-dimensional KV leaves attention computation largely unchanged. We introduce NanoKV, a frozen-backbone framework for generation-aware high-ratio KV compression. NanoKV exploits redundancy across tokens and attention heads by storing shared content in cluster centers and encoding joint head-channel residuals into low-bit coefficients. Crucially, NanoKV uses different fitting objectives for Keys and Values according to their distinct roles in attention: Key decoders are optimized to preserve downstream block outputs, whereas Value encoders and decoders are fitted for faithful reconstruction through the deployed quantization path. Building on the same compressed representation, the optional NanoKV-Fast variant reuses the stored coefficients for mixed-dimensional attention, reducing computation without modifying the video backbone. Across three autoregressive video backbones, NanoKV consistently improves the quality-storage trade-off. On HY-WorldPlay, NanoKV achieves $26.09\times$ effective KV compression, nearly twice that of KVTC, while improving PSNR by 1.0 dB. NanoKV-Fast retains $20.53\times$ effective compression and achieves an approximately $1.15\times$ end-to-end speedup over KVTC.
When many inputs share a budget of repeated attempts by a large language model (LLM), a common rule gives harder inputs more. On public math, code and formal-proof samples, the best published difficulty-proportional rule captures half or more of the oracle gain on benchmarks the model mostly solves, but at most 13% on hard workloads, even given true pass rates. Because the value of one more attempt peaks at intermediate difficulty, the optimal static allocation is inverted-U and gives little to inputs that are hopeless at the current budget. In a strict survey of 221 LLM allocation papers, at least 13 give the hardest inputs the most attempts; their easier settings sit at budgets where this is near-optimal, so the cost appears on harder workloads. Acting on it requires knowing which inputs are solvable, so information, not the policy, drives the gain. With a verifier and a prior from a related model's past runs on the same inputs, Thompson Sampling raises the solved share on the hardest workload from 14% to 27%; without such a prior, no policy we tested beats round-robin. For the policies we tested, the gain vanishes between prior rank correlations of about 0.3 and 0.55, which cheap text-based priors barely reach on the weakest models, and it requires a verifier whose false acceptances are rare relative to the pass rate. On new Qwen2.5 samples the pattern holds, and tempering the prior avoids losses on easy workloads. We turn these results into a decision guide.