AI News Today

arXiv:2606.18807v1 Announce Type: cross Abstract: The field of learning-augmented algorithms has demonstrated that machine-learned predictions can bypass worst-case lower bounds across a wide range of problems. So far, however, the focus has been almost exclusively on polynomial-time algorithms, where predictions improve competitive ratios, approximation guarantees, or running times. In this paper, we raise the question of whether predictions can push the frontier of exact exponential-time algor

arXiv:2606.19023v1 Announce Type: cross Abstract: The growing reliance on pre-trained Machine Learning (ML) models has introduced new attack surfaces. Recent vulnerabilities demonstrate that malicious behavior can be embedded within model artifacts, often bypassing existing defenses. Current model-scanning solutions primarily rely on static, format-specific rules or known attack signatures, which limit their ability to generalize across frameworks and to detect novel exploitation paths. In contr

arXiv:2606.19039v1 Announce Type: cross Abstract: The mismatch between continuous acoustic signals and discrete event-driven processing remains a fundamental bottleneck for neuromorphic speech processing. Current systems typically rely on fixed spike encoders, forcing downstream Spiking Neural Networks (SNNs) to compensate for non-adaptive input representations. To address this, we present a learnable residual speech-to-spike encoder jointly trained end-to-end with a Recurrent Leaky Integrate-an

arXiv:2606.19057v1 Announce Type: cross Abstract: Large Language Models (LLMs) are increasingly used as judges for scalable evaluation, yet such LLM--as--a--Judge systems exhibit systematic biases that are decoupled from semantic quality, most notably verbosity bias. Meanwhile, human supervision is costly and typically selective, yielding reliable positive judgments but leaving most outputs unlabelled and potentially mixed in quality. We formulate LLM evaluation under selective human supervision

arXiv:2510.09905v2 Announce Type: replace-cross Abstract: When an AI assistant remembers that Sarah is a single mother working two jobs, does it interpret her stress differently than if she were a wealthy executive? As personalized AI systems increasingly incorporate long-term user memory, understanding how this memory shapes emotional reasoning is critical. We investigate how user memory affects emotional intelligence in large language models (LLMs) by evaluating 15 models on human-validated em

arXiv:2511.00802v2 Announce Type: replace-cross Abstract: With data-driven development now widely adopted, online A/B testing is an established method for measuring the effects of new technologies. However, deploying online experiments demands resources for design, implementation, and deployment, and may negatively impact users (e.g., unsafe or unethical outcomes) while requiring weeks of data collection. To address this, the growing research area of off-policy evaluation (OPE), or offline A/B t

arXiv:2601.12805v4 Announce Type: replace-cross Abstract: Large language models (LLMs) have shown growing promise in biomedical research, particularly for knowledge-driven interpretation tasks. However, their ability to reliably reason from gene-level knowledge to functional understanding, a core requirement for knowledge-enhanced cell atlas interpretation, remains largely underexplored. To address this gap, we introduce SciHorizon-GENE, a large-scale gene-centric benchmark constructed from auth

arXiv:2602.00161v2 Announce Type: replace-cross Abstract: In this paper, we formulate the compression of large language models (LLMs) by optimally deleting transformer blocks (``block removal'') as a constrained binary optimization (CBO) problem that can be mapped to a physical system (Ising glass), whose energies are a strong proxy for downstream model performance. This formulation enables an efficient ranking of a large number of candidate block-removal configurations yielding many high-qualit

arXiv:2603.06310v2 Announce Type: replace-cross Abstract: Speech foundation models struggle with low-resource Pacific Indigenous languages because of severe data scarcity. Furthermore, full fine-tuning risks catastrophic forgetting. To address this gap, we present an empirical study adapting models to real-world Pacific datasets. We investigate the impact of data volume, adaptation strategies, and representational drift on speech foundation models for various Pacific languages. Additionally, we

arXiv:2605.16385v3 Announce Type: replace-cross Abstract: Geometric problem solving, as a typical multimodal reasoning problem, has attracted much attention and made great progress recently, however most of works focus on plane geometry while usually fail in solid geometry due to 3D spatial diagrams and complex reasoning. To bridge this gap, we introduce Hilbert-Geo, the first unified formal language framework for solid geometry, including an extensive predicate library and a dedicated theorem b

arXiv:2605.29676v2 Announce Type: replace-cross Abstract: Large language models in Agentic AI systems consume tool schemas and execution results and emit tool invocations as structured data. The default language for that exchange, JSON, was designed for application-to-application interchange rather than token efficiency, so its structural elements impose substantial token overhead. Recent work proposes token-optimized alternatives such as TOON (Token-Oriented Object Notation) and TRON (Token Red

arXiv:2606.01249v3 Announce Type: replace-cross Abstract: On-Policy Distillation (OPD) is a fundamental technique for efficient post-training of large language models (LLMs), with broad applications in agent learning, multi-task enhancement, and model compression. However, OPD training becomes unstable when the teacher and student distributions differ substantially, as teacher supervision on student-generated tokens may yield unreliable policy gradients and even cause optimization failure. This

arXiv:2606.07591v3 Announce Type: replace-cross Abstract: AI coding agents are increasingly used for scientific work, but their end-to-end autonomous research capability remains difficult to verify. We present ResearchClawBench, a benchmark for evaluating autonomous scientific research across 40 tasks from 10 scientific domains. Each task is grounded in a real published paper, provides related literature and raw data, and hides the target paper during evaluation. Expert-curated multimodal rubric

OpenAI's LifeSciBench evaluates whether frontier AI can handle real life-science research across 750 expert-authored tasks, seven workflows, and seven biological domains. Built by 173 PhD scientists with 19,020 rubric criteria, it grades reasoning and decisions, not just recall. The best model, GPT-Rosalind, passes 36.1%, leaving large headroom on artifacts, exact outputs, and operational calls. The post OpenAI Releases LifeSciBench, a 750-Task Benchmark Grading AI Models on Real Life-Science Re

<!-- SC_OFF --><div class="md"><h1>🏇 Open-Source Hong Kong Horse Racing ML Pipeline — Feedback Welcome</h1> <p>Hi everyone,</p> <p>I've been working on an open-source horse racing prediction project focused on <strong>Hong Kong Jockey Club (HKJC)</strong> data.</p> <ul> <li>📦 <strong>Repo:</strong> <a href="https://github.com/catowabisabi/horse-racing-model-training">catowabisabi/horse-racing-model-training</a></li> <li>🌐 <strong>Live Dashboard:</strong> <a href="https://catowabisabi.github.i

In this tutorial, we use NVIDIA SkillSpector to evaluate AI skills for security risks before deployment. We build a corpus of benign and deliberately vulnerable skills, then scan them through SkillSpector's programmatic LangGraph workflow. We organize the risk scores and findings with pandas, then visualize severity and category distributions. We export results in SARIF format, register a custom analyzer, and optionally apply an LLM-based semantic pass. The post NVIDIA SkillSpector Guide: Scanni

<!-- SC_OFF --><div class="md"><p>I graduated with my bachelor's in a top 3 CS program and have had a rough recruiting season. I received a full time offer as AI Product Engineer at a tax software company, where they are trying to become more AI native. It's essentially a PM + AI engineering role.</p> <p>Long term I'd love to work at a frontier lab or in a research/more technical role at an AI startup.</p> <p>So, should I take up the offer or pursue my master's at the same school? I am able to d

<!-- SC_OFF --><div class="md"><p>This question is related to topics like language+ models (including multimodal) and things like "circuit" analyses. I think something related might come up in my work (factuality guarantees for model outputs) and I'm trying to orient to the SoTA.</p> <p>I found <a href="https://www.neelnanda.io/mechanistic-interpretability/emergent-pos">this old post</a> on trying to deduce, for instance, whether a Transformer-based model "knows" which word a token is in. Even i

<!-- SC_OFF --><div class="md"><p>I'm not that well versed in ML yet. I know that "Attention is all you need" was based on work that was done with a couple of high end gaming GPUs at the time. I can afford that.</p> <p>Suppose for arguments sake that I have caught up on ML such that I have the competence to recreate state of the art results should I have access to the required hardware, do I still need access to huge amounts of hardware infrastructure to be able to contribute to the field at a f

<!-- SC_OFF --><div class="md"><p>Hi All, I've been running experiments on targeted SFT for specific capability dimensions on a 31B model. After running small training run to prime the model slightly in the direction I want, then ran a judge across 40 domains scoring six independent quality dimensions. One dimension consistently scored weakest across five runs.</p> <p>I am now training contrastive variants from the same checkpoint - examples with that dimension deep vs examples with it deliberat

Vercel has open-sourced eve, an Apache-2.0 agent framework now in public preview. An agent is a directory of files, with durable execution, sandboxes, approvals, connections, channels, and evals built in. Scaffold with npx eve@latest init and deploy unchanged via vercel deploy. The post Vercel Releases Eve: An Open-Source AI Agent Framework Where Each Agent is a Directory of Files Mapped to Capabilities appeared first on MarkTechPost .

Musicians are accustomed to getting paid each time their creative work is used. Across vinyl/CD sales, streams, radio, cover versions, and those numerous niches like karaoke, there are agreements in place about what “use” means. Underlying this is a simple economic principle: The more something is used, the more money it makes. Generative AI has complicated the definition of use . On the one hand, you could argue that the use of a piece of musical training data happens just once, at the point of

<!-- SC_OFF --><div class="md"><p>Hi everyone,</p> <p>I have a fairly weak undergraduate: a 3.3/5 GPA in Computer Engineering from an average Nigerian university. For my Master's, I studied Artificial Intelligence at an average European university, where I finished with an 8/10 GPA.</p> <p>A condensed version of my Master's thesis was recently accepted at ACL 2026, with a meta-review score of 8/10 and a confidence score of 5/5. It's scheduled for presentation next month.</p> <p>I want to pursue

<table> <tr><td> <a href="https://www.reddit.com/r/MachineLearning/comments/1u84mio/nextlatent_prediction_transformers_r/"> <img src="https://preview.redd.it/efm7zazr2t7h1.png?width=140&height=90&auto=webp&s=c1b7070ca3de62bdc276d7a185c72f6737e6f92e" alt="Next-Latent Prediction Transformers [R]" title="Next-Latent Prediction Transformers [R]" /> </a> </td><td> <!-- SC_OFF --><div class="md"><p><a href="https://preview.redd.it/efm7zazr2t7h1.png?width=2950&format=png&auto=webp&s=444dc71b22bca0c499f

OpenAI introduced Deployment Simulation on June 16, 2026. The method replays past conversations through a new candidate model before release. It then grades the completions to estimate deployment-time rates of undesired behavior. We break down how the pipeline works, the reported 1.5x median multiplicative error, and its limits. The post OpenAI’s Deployment Simulation Extends Pre-Deployment Risk Assessment to Agentic Coding Through Simulated Tool Calls appeared first on MarkTechPost .

<!-- SC_OFF --><div class="md"><p>Character AI, founded by former Google/LaMDA developers Noam Shazeer and Daniel De Freitas, proved that text-based character chat can work as a real entertainment category.</p> <p>But the next chapter might not be better text chat. It might be real-time video interaction.</p> <p>Mel AI recently shared a demo of AI character video chat, and the interesting part is the interaction stack: voice, lip sync, facial reactions, and camera-aware responses instead of just

arXiv:2602.06276v2 Announce Type: replace Abstract: We address the problem of training conversion prediction models in advertising domains under privacy constraints, where direct links between ad clicks and conversions are unavailable. Motivated by privacy-preserving browser APIs and the deprecation of third-party cookies, we study a setting where the learner observes a sequence of clicks and a sequence of conversions, but can only link a conversion to a set of candidate clicks (an attribution s

arXiv:2606.17383v1 Announce Type: cross Abstract: Agentic artificial intelligence systems introduce a new class of model risk. Unlike traditional predictive models, autonomous agents continuously acquire information, form beliefs regarding latent states of the environment, generate forecasts, select actions, and adapt their behavior over time. Existing validation methodologies focus primarily on predictive accuracy and therefore provide limited insight into the quality of the underlying decision

arXiv:2606.17386v1 Announce Type: cross Abstract: End-to-end autonomous driving has achieved state-of-the-art performance on benchmarks and real-world deployments. Its standard training recipe, however, is expensive across all stages: collecting and labeling millions of driving frames is costly, and closed-loop RL on images is bottlenecked by the per-step cost of photorealistic rendering plus a forward pass through a large vision backbone. Self-play in vectorized simulators changes the economics

arXiv:2606.17389v1 Announce Type: cross Abstract: Multimodal Foundation Models are increasingly used as reasoning agents, making reliability, knowing when a model may hallucinate, critical. A common intuition, which we call the Attention-Confidence Assumption, holds that reliability follows from "structural" visual perception: tight attention on relevant regions should signal a trustworthy answer, while scattered attention signals confusion. We challenge this through the VLM Reliability Probe (V

arXiv:2606.17209v1 Announce Type: new Abstract: Test-time scaling for agentic search typically increases depth (i.e., more turns and tokens per trajectory) or breadth (i.e., more parallel rollouts). Here we focus on breadth scaling, showing that standard parallel sampling yields diminishing returns, tracing this to query redundancy at the first turn. When models issue similar first queries across rollouts, the threads retrieve overlapping evidence, and subsequent turns are conditioned on this sh

arXiv:2606.17220v1 Announce Type: new Abstract: Legal case retrieval remains challenging due to the complexity of legal language and the need for precise lexical alignment between queries and relevant cases. Although dense retrieval models have achieved notable progress, empirical studies show that BM25 continues to serve as a strong baseline in this domain. It motivates us to propose a self-evolving framework for rule-driven query rewriting that enhances BM25 without any parameter training. The

arXiv:2606.17269v1 Announce Type: new Abstract: In skill-constrained production-inventory systems, the qualified human capacity available tomorrow depends on training decisions made today: production requires certified workers, certifications decay unless maintained, and training consumes the same scarce worker hours that production needs now. We study a closed-loop skill-constrained model predictive controller that, at every shift, solves a finite-horizon mixed-integer program over production,

arXiv:2606.17289v1 Announce Type: new Abstract: AI systems based on artificial neural networks are being developed with aspirations of pushing the boundary of human mathematical knowledge. A key question for these systems is how much they can reach beyond their training data. Mathematical discovery requires a strong form of out of distribution generalization; the ability to hypothesize genuinely new - and potentially logically more powerful - mathematical structures. It has been hypothesized tha

arXiv:2606.17312v1 Announce Type: new Abstract: Large language models can arrive at the same answer through reasoning paths that are unstable, contradictory, or difficult to rank consistently -- a failure mode especially prevalent in multi-step deductive reasoning. Existing methods assess reliability primarily through output dispersion -- measuring how much sampled answers differ -- but this discards a complementary signal: whether the model can consistently rank competing reasoning candidates.

arXiv:2606.17328v1 Announce Type: new Abstract: LLM agents increasingly maintain long-term memory of user facts across sessions. Yet such memory is usually evaluated by aggregating accuracy over question rows or episodes. Because this approach scores question rows independently, even when several questions probe the same fact, it cannot show how that fact behaves as conditions change. We introduce MemTrace, a benchmark whose unit of measurement is the knowledge point: a single typed fact about t

arXiv:2606.17339v1 Announce Type: new Abstract: Speech offers a uniquely informative window into health by simultaneously engaging neurological, motor, respiratory, and vocal systems. Current clinical speech AI methods have largely progressed through isolated condition-specific studies, making results difficult to compare and generalization difficult to assess. We introduce SpeechDx, a large-scale benchmark for clinical speech AI spanning 12 datasets and 27 tasks across diverse health conditions

arXiv:2606.17368v1 Announce Type: new Abstract: Large language models have accelerated the transition from passive conversational assistants to autonomous agents that can understand goals, plan actions, invoke tools, and execute multi-step tasks. Yet the capability of a single agent remains constrained by its local data, tool permissions, runtime environment, and governance boundary. This paper studies distributed general-purpose agent networks: open peer-to-peer networks in which heterogeneous

arXiv:2606.17405v1 Announce Type: new Abstract: Clinical decision support AI systems (CDSASs) must adapt to evolving patient conditions in real-time while adhering to strict safety constraints. We present an online adaptive framework that integrates Treatment Effect (TE) estimation to quantify clinical benefits, a patient Digital Twin (DT) to simulate treatment trajectories, and Reinforcement Learning (RL) for sequential decision-making. The AI system is initially trained on historical medical r

arXiv:2606.17443v1 Announce Type: new Abstract: Large language models (LLMs) are becoming a major way for consumers to find products, but we do not yet understand how brands compete in this new channel. We study brand dynamics in LLM recommendations using skincare products -- a category where consumers cannot easily judge quality before buying and must rely on brand reputation -- across three commercial LLMs (GPT-4o-mini, Claude Sonnet, Gemini 3 Flash), with a robustness check on search goods. I