The Archive

As large language models are increasingly deployed for clinical text, ensuring they can reliably signal their own uncertainty becomes critical. Most existing uncertainty quantification (UQ) methods are designed for open-domain generation and cannot localize uncertainty at the token or span level in long clinical text. We propose Reverse Probing, the first UQ framework specialized for clinical summarization, which estimates token-level uncertainty directly from pre-existing labeled summaries. Rather than sampling new outputs, Reverse Probing treats the text as a probe into the model's internal...

NVIDIA RTX provides game developers with direct paths to AI-driven characters, frame generation, and ray-traced rendering. This post walks through a meaningful... NVIDIA RTX provides game developers with direct paths to AI-driven characters, frame generation, and ray-traced rendering. This post walks through a meaningful set of recent updates across the RTX ecosystem. NVIDIA ACE expands its multilingual AI character capabilities, making it easier to ship conversational NPCs. NVIDIA DLSS 4.5 arrives as an Unreal Engine (UE) plugin… Source

Tabular data in knowledge-rich domains often carries a latent prior in the form of Boolean implication relationships (BIRs) between pairs of features. We mine such relationships with a sparse-exception binomial test. The mined implications form a typed directed graph, equivalent to a propositional rule base of 2-literal clauses. We encode this graph as the connectivity of a layered neural network, called BIRDNet, in which each hidden unit corresponds to one mined rule and binds only to its two features. We show two consequences of this design: First, the architecture is sparse by construction...

Every time I watch someone use Claude Code on a real codebase, the same thing happens. It rewrites a module that three other modules depend on without any awareness of coupling. It just reads the file, makes changes, moves on It reads files one at a time without any map. Doesn't know which files are coupled. Doesn't know who owns what. Doesn't know why that weird pattern in the auth module exists on purpose. I've been building an open source MCP layer to fix this called repowise. Self-hosted, pip install, AGPL-3.0. Five context layers that sit between your codebase and the model: Graph - ...

A general-purpose language model that answers a harmful question returns text; a coding model that complies with a malicious request can return a working weapon -- a keylogger, a ransomware stub, an exploit that runs as written. This asymmetry in the severity of a single act of compliance implies coding-specialized models should clear a higher refusal bar than general-purpose chat models, not a lower one, yet the field cannot presently tell whether they do. Refusal benchmarks for malicious code are fragmented: they mix requests for executable software (ready-to-run weapons) with requests for ...

Product images strongly influence consumer decision-making in online marketplaces. Empowered by multimodal contrastive learning, generative AI can output images that closely align with text prompts. Yet existing generative AI models do not directly optimize marketplace performance. This is a critical gap, since semantic alignment alone does not guarantee that an image will sell. To address this limitation, we propose a \textit{utility-aware multimodal contrastive learning} framework that incorporates consumer demand into a novel Utility-Aware InfoNCE loss. Optimizing this utility-aware object...

Memory is essential for enabling large language models to support long-horizon reasoning, yet existing memory systems remain unreliable and difficult to debug. Tracing memory's dynamic evolution is crucial to understand how information is synthesized, propagated, or corrupted over time. In this work, we study the new problem of error tracing and attribution in LLM memory systems. We propose a novel framework that transforms memory pipelines into executable memory evolution graphs, enabling fine-grained tracing of operational information flow. We then construct MemTraceBench, a benchmark colle...

Designing a transit network requires many sequential route extension decisions, but their quality is often visible only after the full network is assembled. This delayed-feedback challenge lies at the heart of the Transit Route Network Design Problem (TRNDP), where route interactions can be deceptive: an extension that appears useful locally can create transfer bottlenecks, produce redundant overlap, or reduce overall throughput. To guide route construction under delayed simulator feedback, we introduce AlphaTransit, a search-based planning framework for cityscale bus network design. AlphaTra...

Robustness of neural networks is commonly quantified via local or global Lipschitz constants. However, Lipschitz continuity can be overly coarse or overly restrictive as global robustness measure, failing to capture nuanced, data-dependent behavior. We propose a data-driven, architecture-agnostic framework based on the discrete modulus of continuity (DMOC), a non linear generalization of Lipschitz continuity that provides a finer notion of robustness. Unlike many existing approaches, DMOC does not require access to model internals and instead evaluates regularity relative to the data distribu...

We discover that VLA architectures fail in fundamentally different, predictable ways at the motor-command level. Running VQ-BeT, Diffusion Policy, and ACT on identical evaluation protocols (n=450 episodes across PushT and ALOHA 14-DOF bimanual manipulation), we find: (1) direction reversal rate is a universal failure predictor across all three architectures (AUROC=0.93, 0.79, 0.91; p<0.001); (2) jerk monitoring is predictive only for discrete-token architectures, following a discrete-to-continuous gradient (0.88, 0.69, 0.41); (3) velocity violations alone are non-predictive everywhere (AUROC ...

Representation intervention has emerged as a promising paradigm for aligning large language models toward desired behaviors without modifying model weights. Existing methods typically apply a fixed intervention uniformly across all inputs. However, we find that the appropriate intervention direction and strength vary substantially across samples, and such indiscriminate intervention leads to degradation of general capabilities on benign inputs. To address these challenges, we propose Multi-Adapter Representation Interventions via Energy Calibration (MARI). Specifically, we introduce a competi...

Are LLM-based search agents genuinely searching, or using the web to verify what they already know? We study this question on BrowseComp with three diagnostics. Our analysis reveals Intrinsic Knowledge Dependence (IKD): even with tool access, agents often rely on intrinsic knowledge -- information encoded in the model before retrieval -- rather than on external evidence. Agents answer up to 44.5% of BrowseComp questions without tools, generate more than half of their search queries from internally produced hypotheses rather than retrieved leads, and perform worse than closed-book baselines wh...

Modern datacenter RDMA is bottlenecked at the network interface, not the wire. A NIC running RoCE or InfiniBand holds per-connection state for every (application, remote-endpoint) pair - hundreds of megabytes at 1024-application fanout - and pays a four-traversal PCIe round trip on a 64-byte operation, inflating latency an order of magnitude beyond the wire. Both follow from the Queue Pair over PCIe abstraction RDMA inherits from InfiniBand. Huawei's Unified Bus (UB), a public 2025 specification, changes the abstraction: it decouples per-application endpoint state from per-host transport stat...

Anthropic are strongly rumored to be about to have their first profitable quarter. Stories are circulating of companies surprised at how expensive their LLM bills are becoming from usage by their staff. I think this is because OpenAI and Anthropic have both found product-market fit. Enterprise customers are now paying API prices I think they've found product-market fit And they're ramping up The AI-failure stories around this are pretty thin We also know the labs are spending a lot API revenue is becoming less important April is a new inflection point Enterprise customers are now paying API p...

An Initial Public Offering (IPO) filing is a document released when a private firm goes public, allowing individual (retail) investors to purchase its shares. These filings describe a firm's business, financials, and risks and are long, multimodal documents with narrative text and images. Despite their importance to financial markets, there is no large-scale, standardized dataset or benchmark for studying IPO filings with modern language and multimodal models. These documents pose significant challenges: filings frequently exceed 500,000 tokens and lack consistent structural organization. We ...

Context compression aims to shorten long context inputs with minimal information loss for LLM inference acceleration. While existing methods have shown promise, they typically rely on complex compression modules or compression-specific training, leaving the intrinsic capabilities of LLMs underexplored. In contrast, this work reveals that a thinking model itself can naturally compress long contexts by organizing task-relevant information. We thus derive Thinking as Compression (TaC), a new compression paradigm that treats thinking itself as compressed context. Without relying on specific dedic...

Hi all, Sorry for going missing — we’ve been collecting a larger, higher-quality set of more complex tasks. We’re excited to share a major leaderboard update covering the past three months. We’ve updated the **SWE-rebench leaderboard** with **110 fresh Python tasks** from GitHub PRs created in **March, April, and part of May**. The setup follows the standard SWE-bench format: models read real PR issues, edit code, run tests, and must make the full test suite pass. This time, instead of our usual monthly updates with a smaller number of tasks, we collected a larger batch so we could evalua...

Last month NVIDIA released [SOL-ExecBench](https://research.nvidia.com/benchmarks/sol-execbench), a new benchmark of 235 production CUDA kernels lifted from DeepSeek, Qwen, Gemma, and Kimi. We took several top-ranked AI-generated submissions and tried using them in production workloads. Many of them broke, sometimes in surprising ways. One of those kernels is the fused embedding-gradient + RMSNorm backward pass, which runs at the end of every transformer training step. We took the fastest submission on the benchmark for it, and dropped it into the training loop of a small transformer. The ke...

The distortion-perception (D-P) tradeoff is a fundamental phenomenon of Bayesian inverse problems, which characterizes the inherent tension between distortion performance and perceptual quality. Enabling flexible traversal of the D-P tradeoff at inference time is crucial for practical applications. Despite the recent success of diffusion models in zero-shot inverse problem solving, efficient and principled strategies for D-P traversal in diffusion-based inverse algorithms remain inadequately characterized. In this paper, we propose a stage-wise framework for realizing D-P traversal using a si...

Large language models (LLMs) are increasingly used for the automatic evaluation of generated text, yet most prior work focuses on English. Despite the growing demand for multilingual evaluation, extending LLM-based evaluators to multilingual settings remains challenging, particularly for low-resource languages and scenarios where in-domain data is scarce. This work explores several strategies for developing multilingual LLMs-as-a-judge, considering whether in-domain data is available for fine-tuning or not. We systematically analyze English, Spanish, and Basque, representing high-, mid-, and ...

Critical decision-making in socially consequential spaces is increasingly involving AI systems at varying capacities. Yet, despite the ubiquity of autonomous systems, most approaches to handling autonomous moral decision-making resort to scalar or binary judgments. These methods are insufficient for acceptable moral reasoning, as they provide little explanation, leaving out imperative contextual and theoretical information that must be included to support accountability. For this, we propose a framework to model moral reasoning as a distribution over normative ethical theories or ethical plur...

In-context learning (ICL) derives its power from enabling Large Language Models to adapt to new tasks via prompt-based reasoning alone, entirely bypassing the need for parameter updates. Existing theories primarily study ICL in single-task settings, while real-world prompts often contain sequences of heterogeneous tasks, leaving a gap in understanding whether Large Language Models implicitly perform continual learning during inference. To bridge this gap, we propose the first theoretical framework for in-context continual learning, modeling how a pretrained Transformer processes multiple sequ...

Most existing expressivity theories for neural networks assume exact real arithmetic, whereas practical neural networks are executed under finite-precision floating-point arithmetic with implementation-dependent execution semantics. Recent works have begun studying the expressive power of floating-point neural networks, but existing results are limited to highly restricted activation functions and idealized assumptions such as fixed left-to-right reduction orders and correctly rounded activation implementations. In this work, we study the expressive power of floating-point neural networks und...

Baldwinian and Lamarckian evolution have existed for a long time in evolutionary algorithms (EAs) without ever dominating the academic literature or practical applications. In this work, we use modern empirical and theoretical methods to revisit Lamarckian and Baldwinian evolution and rigorously compare them with the generic Darwinian evolution. On the empirical side, we run a comprehensive suite of experiments on graphs from six different datasets from the recent GraphBench benchmark on Maximum Independent Set and Maximum Cut problems. Our results show that Baldwinian and Lamarckian evolutio...

The GSM-Symbolic benchmark (Mirzadeh et al., 2025) reported consistent performance drops across 25 Large Language Models (LLMs) when tested on template-generated variants of GSM8K problems, concluding that the models lack genuine reasoning capabilities. We argue that this conclusion rests on shaky statistical ground. Re-evaluating 20 open-weight models using Generalised Linear Mixed Models with per-question random effects, we find that only half exhibit statistically significant performance changes under the original prompt format. Moreover, we identify a previously unacknowledged factor: the...

Large language models increasingly rely on either reinforcement learning or multi-agent prompting to improve reasoning, yet these two paradigms remain difficult to combine. Directly applying single-agent reinforcement learning to multi-turn multi-agent systems faces following dilemmas: i) Sparse rewards, role-level free-riding and excessive training overhead. ii) Agents only imitate to collaborate. iii) Fixed collaboration protocol falls into oscillating local optimum. We introduce TRACER, a turn-level reinforcement framework for cooperative multi-LLM reasoning. TRACER separates collaborative...

Speckle tracking echocardiography (STE) is the clinical standard for myocardial strain estimation. Despite good performance on global strain (GLS), its accuracy for regional strain remains limited, even though this biomarker is highly relevant for early diagnosis and the characterization of subtle abnormalities. from clinical data. Deep learning is a promising alternative, but its development is constrained by the lack of reliable motion references. Existing solutions rely either on STE-derived labels or on simulations generated by physics-based models, but these synthetic sequences still hav...

Backpropagation is the core learning mechanism underlying deep learning. However, whether and how this algorithm is implemented in the brain remains highly debated. In particular, while forward activations of pretrained models reliably map onto the cortical hierarchy of visual processing, it is unknown whether backpropagated gradients exhibit a similar correspondence. Here, we address this question using functional magnetic resonance imaging (fMRI) and magnetoencephalography (MEG) recordings of human brain responses to natural images. For this, we extend standard encoding analyses of forward ...

Many applications in quantum simulation, quantum chemistry, and quantum machine learning require not a single quantum state but an ensemble of states characterizing the heterogeneity of a target system. Preparing such ensembles state-by-state is prohibitive in both variational and fault-tolerant settings, motivating a generative-modeling approach. We introduce latent-conditioned parameterized quantum circuits (LPQCs), a hybrid quantum-classical framework in which classical neural networks map a latent variable sampled from a prior distribution to the parameters of a parameterized quantum circ...

Reduced-order models (ROMs) can accelerate high-dimensional dynamical simulations, but their accuracy often deteriorates when online dynamics leave the regime represented by offline training data. We develop a projection-based adaptive ROM framework based on incremental singular value decomposition (iSVD), in which occasional full-order operator evaluations provide correction snapshots for online basis updates. The intrusive ROMs considered here are fully parameterized by the basis, so each update naturally propagates to reduced operators and hyper-reduction machinery. Through its evolving si...