The Archive

Aligning text-to-image flow matching models with human preferences via direct reward backpropagation is sample-efficient but hampered by two well-known pathologies: activations cannot be stored across the full sampling trajectory at modern model scale, and chained Jacobian products across steps inflate the reward gradient as it travels back to early indices. Connector-based methods, such as LeapAlign, address these issues by replacing the full backward trajectory with a short pinned path, highlighting a useful decoupling between sampling and optimization. However, the quality of the resulting...

Federated learning (FL) research often begins with a deceptively simple question: What should we try next? A new aggregation rule, a FedProx coefficient, a... Federated learning (FL) research often begins with a deceptively simple question: What should we try next? A new aggregation rule, a FedProx coefficient, a server optimizer setting, a SCAFFOLD variant, or a model architecture tweak may all look promising before an experiment starts. After the run finishes, the harder questions begin: Did the change actually improve the metric? Source

With the widespread deployment of Multimodal Large Language Models (MLLMs) in social interaction, understanding and controlling their behavior under complex personality conditions is essential. This paper introduces explicit personality conditioning and establishes a systematic evaluation framework encompassing single-personality induction, multi-personality induction, and personality switching. Experiments show that personality induction improves image captioning performance but can impair performance on tasks requiring precise reasoning, such as visual question answering (VQA). Balancing an...

Recent advances in reasoning and tool-calling capabilities of large language models (LLMs) have enabled increasingly capable agentic systems. However, existing benchmarks remain limited in task complexity, realism, and domain diversity, and often fail to capture interactions that span multiple domains, limiting their ability to evaluate agents in realistic multi-step settings that require sustained reasoning and coordination. To address these limitations, we introduce T1-Bench, a high-fidelity, comprehensive benchmark for evaluating agentic systems in realistic customer-facing, multi-domain e...

Neural population activity models can recover rich temporal structure from binned spikes, but their read-in and readout layers often remain tied to a fixed set of recorded neurons. This coupling limits reuse in long-term brain-computer interfaces, where recorded neuron identities, counts, and response statistics can change across days. We introduce GRAFT, a Transformer-based neural population activity model that separates reusable temporal dynamics from a recalibratable neuron interface. The neuron interface controls how recorded neurons enter and leave the shared backbone, and auxiliary gain...

AI control protocols oversee untrusted models by monitoring their actions and modifying potentially unsafe steps, often using a trusted model. This partially tampers with the untrusted model's trajectory. If the trusted model detects such an intervention, it may infer properties of the monitor and adapt to evade control. We introduce \textbf{CIAware-Bench}, a benchmark for measuring \textbf{c}ontrol \textbf{i}ntervention (CI) awareness across frontier models. CIAware-Bench tests whether models can distinguish their own trajectories from those modified by a control intervention. The benchmark ...

Despite its importance to applications in protein design, predicting protein properties like binding affinity and thermostability from sparse experimental data remains a significant challenge. Accordingly, we introduce a class of sequence kernels that exploit evolutionary substitution matrices as well as local linearity and demonstrate that the resulting Gaussian processes provide data-efficient models of protein property landscapes, frequently outperforming alternatives that rely on foundation model embeddings. Furthermore--by learning what are in effect structure-aware substitution matrices...

Chain-of-thought (CoT) supervised fine-tuning (SFT) is widely adopted to improve reasoning ability, yet we find that it systematically degrades long-context recall in hybrid linear-attention models. Across architectures including HypeNet and Jet-Nemotron, retrieval performance on Needle-In-A-Haystack (NIAH) deteriorates substantially after CoT-SFT, and the degradation becomes more severe under harder retrieval settings and longer context windows. For example, HypeNet-9B on NIAH-S2@256K decreases from $67.2\%$ to $9.4\%$. We attribute this to CoT-SFT biasing attention gradients toward short-ra...

Instruction-tuned LLMs are increasingly converted into reasoning models through post-training to improve multi-step task performance. This conversion is usually optimized for reasoning accuracy, without explicitly preserving the alignment behavior of the instruction-tuned model, such as safe refusal, bias avoidance, and privacy protection. We ask: does this conversion preserve alignment? We study this question through a trustworthiness audit and find that it is not behavior-preserving by default. For a systematic analysis, we compare reasoning models produced via supervised fine-tuning, RL-ba...

Apple’s feature showcase at WWDC 2026 didn’t flag which if these “photographs” are real or created with its new AI fakery. | Images by Apple / compiled by The Verge Apple used to question whether generative AI-powered editing features were worth the risk of distorting our perceptions of the world. Now it seems Apple no longer believes that photos should accurately capture reality. At WWDC 2026, the company announced a host of new AI-powered photo editing tools. They give users effortless powers of manipulating images that Apple still refers to as "photos." Two years ago, Apple launched Clean ...

Reusing a held-out benchmark adaptively should, in principle, invite overfitting. Yet benchmark-driven machine learning (ML) has produced surprisingly little overfitting in practice. An attractive hypothesis is that successful ML strategies are highly compressible. We study this in the setting of LLM-driven research agents, where the hypothesis becomes directly testable via two complementary information bottlenecks. In \emph{output compression}, an exploration agent adaptively searches for high-performance models using a validation set, and we test whether a fresh ``reproducer agent'' can rep...

Conformal predictive systems (CPS) output calibrated bands of CDFs under exchangeability. We extend generalized CPS to non-exchangeable settings by encoding distributional shifts through observation-specific permutation weights. This yields shift-aware predictive systems that remain valid whenever the test point is, conditionally on the unordered sample, a weighted draw from the observed atoms. Since such weights are typically estimated, we introduce weight-uncertainty boxes and construct robust CPS envelopes with finite-sample or asymptotic confidence guarantees. We derive efficient computat...

Recent years have witnessed the rapid evolution of AI agents toward handling increasingly complex, real-world tasks. However, existing benchmarks rarely evaluate whether agents can operate graphical user interfaces to complete long-horizon, high-value professional workflows across diverse domains. Current GUI benchmarks still predominantly focus on general-purpose software, relatively simple applications, and short-horizon tasks, leaving it largely unknown whether modern agents can follow user instructions to autonomously operate domain-specific professional software and accomplish economical...

With SpaceX, Anthropic, and OpenAI all eyeing massive public debuts, the tech industry may soon have a new class of corporate overlords — and a new acronym to match. Say goodbye to FAANG and hello to MANGOS.

Recent efforts to extend large language models (LLMs) to speech inputs typically rely on cascaded ASR-LLM pipelines, end-to-end speech-language models, or bridge/distillation-based adaptation. While these routes respectively reuse strong pretrained components, enable native speech-language interaction, or offer lightweight adaptation, they often suffer from transcript-interface latency, costly multimodal training, or sequential speech-language coupling. To address these limitations, we present AuRA, a method that distills audio encoding capability into the LLM. Specifically, AuRA feeds the sa...

Recent work has demonstrated that online reinforcement learning (RL) can substantially improve the quality and alignment of flow matching models for image and video generation. Methods such as Flow-GRPO and CPS cast the denoising process as a Markov Decision Process and apply PPO-style ratio clipping to enforce a trust region. However, we argue that ratio clipping is structurally ill-suited for flow models: the probability ratio between new and old policies is a noisy, single-sample estimate of the true policy divergence, leading to over-constraining in some regions of the trajectory and unde...

Sequential recommendation aims to predict users' next interaction with items by analyzing their historical behavior. However, the limited quality of item representations remains a critical bottleneck. While pre-trained large language models (LLMs) can provide rich semantic representations, existing approaches only rely on static encoding of fixed attributes, overlooking the crucial role of target audiences in defining item identity. Moreover, the semantic space struggles to reflect actual user behavior, resulting in a significant gap between semantic representations and behavioral patterns. T...

Learning-based motion planners, despite recent progress, often suffer from temporal inconsistency. Small perturbations across frames can accumulate into unstable trajectories, degrading comfort and safety in closed-loop driving. Several methods attempt to inject history as a static conditioning signal to stabilize outputs, only to induce the planner to copy historical patterns instead of adapting to environment contexts. To address this limitation, we propose Diffusion Forcing Planner (DFP), a diffusion-based planning framework driven by history-guided control. Specifically, DFP decomposes th...

Measuring subjective constructs in naturally occurring social media text requires annotation procedures that are theoretically grounded, empirically validated, and transferable to an encoder model for scalable prediction. Using non-English social media posts annotated according to Schwartz's theory of basic human values, we investigate how different LLMs, prompts, and instruction languages operationalize the expression of values in text. We argue that although texts may permit multiple plausible interpretations, theory-based value definitions can constrain interpretations and reduce spurious ...

Airport surface operations increasingly constrain performance at high-throughput hubs. This study examines arrival taxi-in decisions at Hartsfield-Jackson Atlanta International Airport (KATL) and proposes a two-stage, data-driven decision aid that mirrors controller workflow. Stage I predicts the runway exit selected by an arriving aircraft. Stage II predicts whether, given that exit, the aircraft will cross the active departure runway at a designated point or use the end-around taxiway. Models are trained using ASDE-X surface trajectories, aircraft characteristics, ramp destinations, short-h...

We ask whether large language models (LLMs) merely imitate rationales when choosing between two options, or whether their choices reflect a systematic underlying decision structure. Using synthetic binary decision settings in which models choose between profiles defined by graded attributes, we compare the attribute a model says mattered most with the attribute that best explains its choice under a behavioural model fit to prior decisions. The behavioural model predicts held-out choices well, showing that model behaviour is systematically related to the visible attributes rather than being ra...

Tuning controllers for strongly coupled multi-input multi-output (MIMO) industrial processes is hard: decentralized classical auto-tuning ignores loop interaction, and local numerical optimization from natural initializations stalls in the resulting non-convex cost landscape. We ask whether on-premise open-source large language models (LLMs), which keep data on-site and need no plant model, can help. On a single-loop CSTR, classical relay-feedback tuning (IAE 0.106, near the 0.102 optimum) beats an LLM tuner (0.162): for simple loops the LLM adds nothing. The picture inverts on a strongly cou...

Large language models are increasingly used to adapt math word problems for personalized learning at scale, but it remains an open question whether those adaptations are consistent across models, preserve cultural diversity at scale, and reveal which cultural entities models treat as most salient. We analyze how Claude Opus 4, GPT-4.1, and Gemini 2.5 Pro adapt 60 English math word problems into Bengali, Hindi, Punjabi (India), Urdu, Sindhi (Pakistan), Italian, and Sicilian (Italy), a language set spanning the full resource spectrum, from high-resource Italian and Hindi to under-studied Sindhi...

OpenClaw has rapidly emerged as a transformative artificial intelligence (AI) agent framework, and its ability to autonomously execute complex, multi-step tasks has attracted an ever-growing and diverse user base. However, this capability comes with significant risks. While existing research has made important strides in characterizing these threats, such work is predominantly directed at technically sophisticated audiences. It remains largely inaccessible to non-technical users. This demographic now makes up an increasingly large and underserved portion of the community, yet it is these very...

Large language model unlearning aims to suppress designated undesirable knowledge while preserving benign capabilities. Many unlearning objectives focus on suppressing undesired answers, while recent target-guided variants specify replacement behavior but still leave update locality largely unconstrained. This paper introduces \emph{Null-Space Constrained Response-Specified Unlearning} (NSRU), a projection-constrained low-rank framework for controlled LLM unlearning. NSRU uses an explicitly structured safe target response to specify the desired behavior for each forget query, while suppressin...

Gemini 3.5 Live Translate brings near real-time, natural speech translation to Google AI Studio, Google Translate and Google Meet.

Training a speech AI model to correctly recognize or synthesize clinical terminology is surprisingly difficult. Drug names like Acetaminophen, Amlodipine,... Training a speech AI model to correctly recognize or synthesize clinical terminology is surprisingly difficult. Drug names like Acetaminophen, Amlodipine, Cefazolin, and Biktarvy are not part of everyday vocabulary. Procedure names, anatomy terms, and specialty-specific diagnoses introduce the same problem in a different form. Off-the-shelf speech systems can sound fluent and still miss the words… Source

Microsoft AI head Mustafa Suleyman is walking back his statement about AI automating jobs done by white-collar workers, including lawyers, accountants, and project managers. During an episode of Decoder on Monday, Suleyman says he meant AI will help these workers complete tasks, rather than do their jobs: Sending an email, having a conversation with a colleague, putting together a PowerPoint - sub-tasks will increasingly become digitized, automated, and we can basically generate more and more of them. That does not necessarily mean that the role goes away at all. It just means that the work c...

Apple kicked off its annual developer conference with bold promises about AI. The company, CEO Tim Cook said, would be "introducing new technologies and innovations that push the limits on what's possible." But its slew of announcements - centered on a brand-new "Siri AI" - had more to do with catching up. After almost entirely neglecting Siri and punting its AI promises down the road in 2025, Apple went all in on the tech this year. It pitched Siri as an all-encompassing virtual assistant that ties together all your Apple devices, with multimodal features, a dedicated app, an all-in-one AI a...