1. Agents Are Going Async
Async subagents let an agent stay responsive while background workers run long tasks. How the orchestrator, notifications and approvals work in Deep Agents.
AI NEWS JOURNAL
Updated with the news
THE AI NEWS JOURNAL
AI Daily
Daily essentials. Weekly context. Monthly perspective.
Published-news recap
All published reports for this day, updated as new stories go live.
44 events13 sourcesAbout 36 min read
2. AGORA: Reliability-Aware Aggregation of LLM Judges via Item-Response Theory
arXiv:2611.05556v1 Announce Type: new Abstract: LLM-as-a-judge is a common substitute for costly human evaluation, and aggregating several judges is the standard remedy for any single judge's bias. Judge pools mix judges of very different reliability, while the human-labeled sets that validate them hold a few hundred items or fewer. In this regime the per-item mean weights weak judges as heavily as strong ones, and confusion-table models such as Dawid-Skene must estimate a per-judge error table from too few items. We propose AGORA (Aggregation Of Reliability-weighted Assessments), an item-response-theory aggregator that places every item on one latent quality scale and describes each judge by a sensitivity parameter and ordered rating thresholds, so judge reliability is learned from all items jointly and without human labels. Across three benchmarks, AGORA matches or exceeds Dawid-Skene on large evaluation sets, and its advantage widens as the set shrinks toward a few hundred items. It remains accurate as pools grow more uneven, a label-free dispersion statistic indicates in advance when trust weighting helps, and its learned trust weights track per-judge human validity.
3. AtomicFaaS: Offloading Side-Effect Compensation to the Serverless Platform
arXiv:2611.06601v1 Announce Type: new Abstract: Function-as-a-Service (FaaS) simplifies distributed application development by offloading operational concerns to the platform, but developers remain responsible for recovering from failures. Retries require idempotent functions, which means that developers have to write complex, error-prone code that undoes or skips the already completed operations of a failed function. To address this, we propose AtomicFaaS, a platform extension that provides semantic atomicity for the external side effects of a function invocation, whether the function fails or its instance crashes, without developer-written compensation code. AtomicFaaS intercepts outbound calls and records their effects in a fault-tolerant log, from which the platform compensates leaked effects outside the failed instance. Compared with a manually implemented saga, AtomicFaaS eliminates compensation code and leaves no compensatable effects after any tested failure, while adding ~1 ms per intercepted call. Under load, it lowers the sustainable throughput of CPU-bound functions, while its latency overhead becomes negligible once calls to external services dominate the invocation time.
4. Augmentation with Reinforcement Transformation for Low-textured Scientific Sketch Image Retrieval
arXiv:2611.05408v1 Announce Type: new Abstract: Low-textured sketch images are widely used in scientific applications, including wildlife conservation, archaeological illustration, and forensic composites. Learning from and retrieving such images is challenging because they lack detailed surface textures and visual variations commonly exploited by modern deep learning models. This paper proposes Augmentation with Reinforcement Transformation (ART) to fine-tune pretrained models to learn effective representations for accurate sketch image retrieval. Our key observation is that sketches are particularly sensitive to image transformations, such as rotation, flipping, zooming, and stretching. While some transformations preserve their structural integrity, others may introduce distortions that hinder representation learning. Instead of randomly choosing transformations to construct fine-tuning views, ART formulates augmentation selection as a multi-armed bandit (MAB) problem, where transformations are treated as arms and their effectiveness is continuously evaluated through rewards. This enables ART to adaptively construct reward-guided view banks before self-supervised fine-tuning; the subsequent representation loss does not feed back to ART. On Florida manatee sketches provided by Mote Marine Laboratory, ART achieves 71.40% average Top-1 accuracy for transformed-scar retrieval, compared with 68.83% for random augmentation selection, and 55.84% for location-aware sketch retrieval, compared with 18.18% for frozen CLIP.
5. b12131
The update removes the u8path() function and replaces it with generic_string() in the server component, as part of a broader effort to streamline and standardize string handling across platforms.
6. b12132
The latest release of llama.cpp, version b12132, includes the removal of u8path() in favor of generic_string(), along with updated binaries for multiple platforms including macOS, Linux, Windows, Android, and openEuler, supporting various backends like CUDA, Vulkan, ROCm, OpenVINO, and SYCL.
7. b12135
The latest release of llama.cpp (b12135) introduces support for MEAN operations in WebGPU, enhancing performance and functionality across multiple platforms including macOS, Linux, Windows, Android, and iOS. The release includes precompiled binaries for various architectures and backends such as CUDA, Vulkan, ROCm, OpenVINO, SYCL, and zDNN, with specific builds for Apple Silicon, Snapdragon devices, and openEuler systems. Some features remain disabled or are under development.
8. b12136
The update addresses an issue in llama.cpp where the number of GPU layers was incorrectly calculated by not accounting for skipped NextN layers, leading to improper layer distribution between CPU and GPU. The fix ensures that skipped NextN layers are included in the n_gpu_layers count and properly assigned during model loading, resolving a bug where layer 0 remained on the CPU despite offloading all other layers.
9. b12137
A recent update to llama.cpp addresses a critical issue by rejecting non-object session updates in the server, preventing malformed realtime session data from causing uncaught exceptions and server termination. The release includes binaries for multiple platforms including macOS, Linux, Windows, Android, and iOS, with support for various backends like CUDA, Vulkan, ROCm, OpenVINO, and SYCL.
10. b12138
The update introduces optimized Metal kernel usage for matrix multiplication with K values between 32 and 64, and adds a new kernel for BF16 x BF16 matrix multiplication, improving performance on Apple Silicon devices.
11. ClickHouse expands collaboration with Microsoft, bringing Fabric integration, deeper OneLake interoperability, and enterprise deployment flexibility
ClickHouse expands its Microsoft collaboration with a native Fabric workload, OneLake read and write capabilities, and ClickHouse BYOC availability in the Microsoft Marketplace.
12. ClickHouse Workload for Microsoft Fabric: sub-second analytics on OneLake, now in public preview
ClickHouse Workload for Microsoft Fabric brings sub-second analytics to OneLake data through a native Fabric workload, with dedicated ClickHouse Cloud compute for agents, dashboards, and applications.
13. Cognitive Bias Signatures in Search-Equipped LLM Agents: A Stage-Resolved Analysis
arXiv:2611.05720v1 Announce Type: new Abstract: Search can help LLM agents correct mistakes, but their starting conditions may shape which evidence they find and use. Drawing on cognitive-bias paradigms, we develop a controlled search harness to examine result order, records of prior work, and initial hypotheses while holding each question and document collection fixed. Matched controls distinguish order effects from differences in document composition and the effects of failed work from context length. Across 10 models and 1,000 synthetic search worlds, we track queries, retrieved documents, document openings, and final answers. Our comparisons suggest that final answers are more sensitive to initial evidence order during active search than when all evidence is supplied in full. Dead-end histories can reduce strategy switching relative to matched neutral histories, indicating an effect beyond context length. Under wrong hypotheses, stage measurements distinguish reduced retrieval or opening from errors that persist after counter-evidence is opened. Exploratory interventions show that opening counter-evidence more often can coincide with lower accuracy, while accuracy can improve with little change in opening.
14. Configure, Certify, and Operate Kubernetes GPU Clusters with NVIDIA DSX OS
One underperforming GPU can slow an entire distributed training job. The node may pass basic health checks and show normal utilization, leaving operators to...
15. Craft and taste
What does old me think
16. Discarded or Unreachable? Diagnosing and Repairing Spatial Relation Failure in Vision--Language Similarity
arXiv:2611.05565v1 Announce Type: new Abstract: Contrastive vision-language models score left/right relations near chance, but the symptom has two causes that call for opposite repairs: the relation may survive in the pooled embedding yet be unreachable by native image-text similarity, or pooling may discard it while the dense features retain it. On clean two object photographs, across ten encoders, the lateral polarity is linearly decodable from the flip odd component of the pooled vector at 0.77-0.99, yet a search over a million captions leaves nine of the ten at chance under native scoring: an access failure. On controlled C4 orbits, where the relation is the only thing that varies, pooling keeps the arrangement axis at 0.998-1.000 and loses its polarity, which the patch grid retains: a pooling failure. Both sit downstream of the dense features, so one readout covers both. Our zero parameter, query conditioned spatial moment recomputes the geometry from the patch grid and stores it as a vector beside the native embedding. On a 118,287 image index it raises left/right pairwise accuracy against reversed geometry negatives from 0.51-0.53 to 0.81-0.96 and conditional R@1 among 21 category matched candidates from 0.05 to 0.37-0.75, at under a millisecond per batched query, while reproducing the native ranking exactly for queries that name no relation.
17. Faster Retrieval for Better, More Efficient RAG
Maximum inner product search (MIPS) sits at the core of recommendation systems, search, and retrieval-augmented generation (RAG). Given a query vector, MIPS systems search within a database of millions of candidate key vectors for the one that best matches the query. The original data point associated with that best-matching key is then used to handle the query: in RAG, the document that is best suited to answer the question is typically prefixed to the query. As the number of keys grows to millions and vector dimensions reach thousands, the computational cost of running an exhaustive…
18. From Activation to Intervention: Interpreting Sparse Visual Features through Reconstruction
arXiv:2611.05522v1 Announce Type: new Abstract: Sparse autoencoders (SAEs) decompose visual representations into individual features, offering a promising basis for explaining model behavior. Interpreting these features typically relies on images or pixel regions that most strongly activate them. However, this evidence becomes ambiguous when multiple features with different semantics frequently activate in the same locations, such as texture and color in the same patterns. We propose Reconstructive SAE Intervention (RecSAEI), a framework that shifts feature explanation from passive observation to active visual reconstruction. By intervening on representations along SAE decoder directions and reconstructing the visual output, we provide direct evidence of a feature's semantic meaning. We demonstrate that our method successfully disambiguates feature meanings across multiple datasets and vision encoders. Furthermore, our framework grounds [CLS] SAE interventions in a sparse set of patch SAE decoder directions. Beyond stable target response, this attribution also exposes the compositional relationship between elementary visual patterns and high-level semantics. Our results advocate for intervention reconstruction as an essential complementary signal for explaining neural networks.
19. From Tokens to Trajectories: Parameter-Efficient Continuous Adaptation of Unified Multimodal Models
arXiv:2611.05451v1 Announce Type: new Abstract: While discrete tokenization in Unified Multimodal Models (UMMs) establishes strong semantic reasoning, it introduces a strict quantization bottleneck that degrades fine-grained visual details, often resulting in structurally accurate but artifact-heavy images. We introduce FlowShift, a parameter-efficient post-training framework that bypasses this bottleneck by converting pre-trained discrete UMMs into continuous rectified-flow generators. Instead of discarding the backbone's semantic priors, FlowShift replaces discrete visual interfaces with continuous latent projections and applies LoRA to the attention layers. To efficiently boost visual quality without semantic drift, we propose Dual-Objective Velocity Vector Distillation (DOVD), which aligns the model with a high-fidelity continuous teacher and ground-truth continuous trajectories. This approach drastically improves perceptual quality while preserving prompt adherence and reasoning. When applied to leading discrete UMMs, FlowShift significantly improves COCO-30K FID and aesthetic scores, while retaining over 99.5% of multimodal understanding capabilities. Ablations confirm DOVD drives generation quality. Ultimately, FlowShift achieves high-fidelity continuous generation without sacrificing pre-trained semantic grounding.
20. Gemini Omni 1.2 Flash: Fast and cost-effective with longer clips
Google DeepMind has launched Gemini Omni 1.2 Flash, introducing native 15-second video generation, improved reference adherence, stronger consistency across frames, multi-turn conversational editing, more natural speech, and cinematic camera movements. The model is now generally available via the Gemini API and Gemini for business, priced at $0.07 per second for 720p video—30% lower than its predecessor. It supports scalable workflows, including draft generation at lower resolutions, iterative edits, and upscaling to 1080p or 4K. Use cases include product explainers, social ads, virtual try-ons, and branded content, with applications across Google AI Studio, Google Cloud, Google Vids, Flow, the Gemini app, and Google Ads.
21. How Amazon Marketplace evaluates content quality with Amazon SageMaker AI
Learn how Amazon Marketplace evaluates A+ Content quality at scale on Amazon SageMaker AI by rendering each product page as a single snapshot, scoring it with a vision-language model fused with OCR readability signals in a CatBoost classifier, and using SHAP to turn each decision into prioritized, actionable recommendations for selling partners.
22. How to assess your organization’s AI readiness
AI adoption moves fast. Readiness takes longer. Assess 6 dimensions to see where your data foundation limits AI.
23. Lesion Overlap Ranks PET Segmentation Models but Does Not Certify Patient-Level Tumor Burden
arXiv:2611.05564v1 Announce Type: new Abstract: Metabolic tumor volume (MTV) summarizes whole-body PET disease burden for prognostic stratification and treatment assessment. Automated MTV is useful only if its estimate is reliable for the individual patient, yet segmentation models are usually evaluated with overlap metrics such as Dice rather than burden error. We examined whether overlap metrics identify unreliable MTV estimates using 1,014 autoPET FDG PET/CT studies. We evaluated LesionTracer predictions held out from fine-tuning, compared five selected released models on 489 tumor-positive patients, and examined the findings in 597 PSMA studies. Across the five FDG models, ranking by higher mean Dice gave the same order as ranking by lower mean relative MTV error ($\rho=-1.00$, exact $p=0.017$). At the patient level, however, out-of-fold Dice tracked relative MTV error strongly ($\rho=-0.75$), yet 82 of 489 patients (16.8\%) fell on opposite sides of the Dice 0.80 and MTV-error 25\% thresholds. Within the training-exposed released models, correlations ranged from $-0.47$ to $-0.74$ over narrower, high-Dice ranges. Out of fold, opposing false-positive and false-negative volumes cancelled about one third of spatial error in net MTV. In 236 of 489 patients, each of the five models missed at least one lesion, though not necessarily the same lesion. Thus, agreement in average model rankings did not establish the reliability of individual MTV estimates or reveal shared lesion misses. PET segmentation benchmarks should report patient-level MTV error and missed-lesion counts alongside Dice and other overlap metrics.
24. Memory safety for Postgres extensions in C/C++
How ClickHouse’s Postgres extensions handle the memory safety challenges of combining C and C++, from clean language boundaries to isolated helper processes.
25. New NVIDIA Cosmos 3 Open Models for Interactive World Simulation
NVIDIA has introduced four new open-weight checkpoints for its Cosmos 3 model family—Cosmos3-Nano-Transfer-Auto, Cosmos3-Nano-Sim-Auto, Cosmos3-Nano-Sim-Transfer, and Cosmos3-Nano-Sim-Bimanual—enabling interactive world simulation and multi-sensor, multi-view driving scene generation. Each model is a post-trained variant of Cosmos 3 Nano with 16B trainable parameters and is based on OpenMDW1.1, allowing commercial and non-commercial use. The checkpoints support applications in robotics and autonomous driving by generating synchronized camera and LiDAR data, enabling streaming simulations, translating simulator controls into photorealistic video, and predicting visual outcomes from robot actions. They are designed to be adapted using custom data and training recipes, with support for offline data generation, causal distillation, and interactive rollouts. Resources include technical reports, demos, and notebooks for inference and adaptation.
26. Oracle wants to pull stalled AI agents out of demo mode and into production
Companies want AI agents faster than they can build the AI architecture to run them. Most can put together a convincing demo. Far fewer can move agents into production without exposing data, hurting uptime or blowing their budgets. That gap is pushing database providers to rethink what a data platform for agents has to deliver. […] The post Oracle wants to pull stalled AI agents out of demo mode and into production appeared first on SiliconANGLE.
27. Performance Portability of Out-of-Kernel Optimizations on GPUs: A Molecular Docking Case Study
arXiv:2611.06207v1 Announce Type: new Abstract: Achieving performance portability is challenging for irregular scientific applications built as repeated multi-kernel GPU pipelines, because throughput depends on out-of-kernel decisions as much as on kernel efficiency. Using molecular docking as a representative case study, we propose an out-of-kernel tuning methodology that treats batching policy, worker concurrency, and per-worker memory allocation as explicit optimization parameters, replacing backend-dependent occupancy-driven choices with a more portable execution strategy based on explicit resource bounds. We implement this methodology in the muDock mini-app under a manager/worker execution model where multiple host workers share a GPU through independent streams. We then evaluate the methodology on NVIDIA, AMD, and Intel GPUs across CUDA, HIP, and SYCL implementations. Across all evaluated platform/backend pairs, the memory-bounded formulation exceeds the peak throughput of the occupancy-guided baseline. The fully tuned methodology improves the end-to-end throughput on all tested configurations, achieving an average speedup of 1.53x over the baseline.
28. Phonological Knowledge of Chinese Characters in Language Models
arXiv:2611.05803v1 Announce Type: new Abstract: How much phonological knowledge do language models (LMs) possess about Chinese characters, and how is this knowledge represented and used? We introduce PinyinBench, a controlled benchmark covering homophones, phonological features (initials, finals, and tones), and polyphonic characters. We investigate 17 text-only and multimodal LMs through three complementary approaches: behavioral evaluation, layer-wise probing, and activation steering. Behaviorally, models demonstrate substantial knowledge of initials and finals, with a frontier model (DS-v4-flash) scoring 90%+ in homophone judgment, when two characters share no phonological features. However, its accuracy drops sharply to 72% when two characters differ only in lexical tone. Probing reveals that tone is highly decodable from embeddings and improves modestly across layers, whereas segmental information starts less decodable and improves substantially with depth. Activation steering shows that segmental representations can systematically influence generation, while tone exhibits much weaker effects. Furthermore, audio-bearing training improves phonological decodability, including tone, without corresponding improvements in steering effectiveness. Together, these findings reveal distinct representational and functional profiles for segmental and tonal information, with tonal information much harder to access and manipulate, demonstrating the necessity for combined evaluation, probing, and causal intervention for understanding linguistic knowledge in LMs.
29. Physics-Informed Identification of Tensegrity Prestress across Taut-Slack Transitions
arXiv:2611.04007v1 Announce Type: new Abstract: Prestress governs the stability, stiffness, and dynamic behavior of tensegrity structures, but it is not directly observable from sparse structural measurements, and it may decrease during service life. Under strong excitation, the cables may slacken and the dynamics become nonsmooth. Existing identification methods are limited to the taut regime, and physics-informed networks for tensegrities solve only the forward problem of predicting the response from known member parameters. This study formulates the inverse problem. Neural networks are trained on the equation-of-motion residual in constrained Cartesian coordinates, where the mass matrix is constant and the force densities enter linearly. The force densities, damping, and axial stiffness are trainable parameters, and the tension-only cable law is kept. A Fisher-information analysis shows, before inversion, which member parameters a record can determine. The method is tested on simulated planar tensegrity frames with up to five modules and a three-dimensional prism, and on measured records of a biomimetic leg and a full-scale tensegrity simplex. On the one-module frame, the identifiable force densities are recovered within 1.5% from noisy accelerations at half the free coordinates, on records with and without taut-slack transitions. Records with transitions determine members that taut records leave undetermined. On the five-module frame, a 16-cable two-span bridge, the median force-density error rises moderately, to 4.2%. On measured records, an identified spring stiffness is within 0.1% of its independently measured value on average, and the identified simplex prestress at 13 levels agrees with its cable force transducers to 2.9% root-mean-square.
30. Prospective Memory for Proactive LLM Agents
arXiv:2611.05731v1 Announce Type: new Abstract: Most memory systems for LLM agents are retrospective, preserving and retrieving past information to inform current decisions. Proactive agents, however, also require prospective memory: retaining future-directed intentions and realizing them when appropriate opportunities arise, even without explicit requests. Crucially, remembering the right intention does not guarantee acting at the right time: a related observation may call for waiting, while a later cue may make the same intention actionable. Inspired by cognitive theories of human prospective memory, we introduce Pro-Memory, a memory framework that separates maintaining future intentions from deciding when and how to realize them. Pro-Memory preserves pending objectives linked to their supporting evidence, assesses new observations against their original goals and activation conditions, and coordinates appropriate contributions with the agent's ongoing actions. It further updates intention states based on actions actually issued or evidence of resolution, while an experience memory distills feedback from completed episodes into guidance for future decisions. Experiments on two benchmarks spanning delayed-intention tasks and open-ended proactive assistance demonstrate consistent improvements in proactive task performance over representative memory approaches across multiple LLM backbones, while limiting unnecessary interventions.
31. Rewrite, Then Veto: Rewriter-Agnostic Edit Gating for LLM-Based Post-OCR Correction
arXiv:2611.05539v1 Announce Type: new Abstract: LLMs can repair noisy OCR transcripts of historical documents, but they also change text that was already correct. We propose EditVeto, a veto gate placed after a full-context LLM rewrite. The rewrite is character-aligned to the OCR input, its differences are split into blocks, and a 0.6B-parameter gate keeps or reverts each block. An oracle analysis motivates the design: given the true error locations, per-span generation is worse than no correction, whereas vetoing a full rewrite's edits helps. Trained on rewrites by two Qwen3 models and evaluated, frozen, on the eight HIPE-OCRepair-2026 test sets, the gate improves all five rewriters tested, two of them unseen in training: weighted cMER falls by 0.0026-0.0044 and the preference score rises by 0.15-0.25 (paired bootstrap, p ~ 0.0002 each). On published competition outputs it lowers cMER for three of four teams but raises it for the winning run (0.0050 to 0.0065). In a second, disclosed test look, a retrained gate still improves LoRA-tuned rewriters (cMER -0.0011 to -0.0021), and an ensemble of their gated edits reaches a cMER of 0.0069 and a preference score of 0.928, mostly through fine-tuning and pooling rather than gating. Built only on public models and data, it is behind the winner (0.0050) and tied with the second-ranked run (0.0071), both from the only team using domain pre-training, and ahead of all other published runs. A third, secondary look shows the unchanged gate improving zero-shot Qwen3.8-27B (cMER 0.0114 to 0.0108) but only the preference score of its LoRA-tuned version.
32. Same model, better answer: Enterprise data is what turns generic AI into a business tool
Generic AI models can write SQL, but they don’t know a company’s orders, customers or revenue definitions. Now, Oracle Database aims to give AI agents the context to move a stalled order forward. But that shift raises questions about complexity, lock-in and security as AI agents start acting on live transactions. Oracle Corp.’s answer is […] The post Same model, better answer: Enterprise data is what turns generic AI into a business tool appeared first on SiliconANGLE.
33. Show HN: calsdk, one TypeScript API for Google Calendar, Outlook and iCloud
Article URL: https://github.com/vkpdeveloper/calsdk Comments URL: https://news.ycombinator.com/item?id=50383335 Points: 2 # Comments: 0
34. Show HN: Cerea – yet another self-hosted LLM chat front end
Here is Cerea, a self-hostable, open-source (Apache 2.0) chat+agent frontend for LLMs. It works alone with any OpenAI-compatible inference provider, but it's way better when coupled with the model gateway I'm also developing, Pystino (https://github.com/paoloviviani/Pystino).I came to this after struggling to find a hosted chat interface that was truly open source, that I really enjoyed to use, and that was simple enough to be shared and used by family. In particular, I was looking for something that could provide the smooth and integrated user experience of commercial products for the tasks needed by family (e.g., document editing, form filling, notes and knowledge management), but backed by open models hosted wherever I decide.Then it became richer, now supporting: remote coding agents (wrapping Opencode), delegation and coordination between agents, integration between chat/projects and code agents (this one to me is a pain point for Claude, which I use at work), projects, RAG, rich artifacts support, schedules, and more.You can try it yourself, either cloning the full repo and head to Cerea/kit, or just the deploy kit with a script (https://github.com/paoloviviani/Cerea#quick-start).It's just a docker compose with a ./configure script that helps to generate a sensible .env for the specific scenario. It works in the browser and as a PWA on mobile.I hope you'll enjoy, feel free to reach out for questions. Comments URL: https://news.ycombinator.com/item?id=50384502 Points: 1 # Comments: 0
35. Show HN: Flawless UI/UX – A design skill and inspection CLI for coding agents
Article URL: https://github.com/smokeeaasd/flawless-ui-ux Comments URL: https://news.ycombinator.com/item?id=50381846 Points: 1 # Comments: 0
36. Show HN: Ls-service – minimal, namespaced, and typed localStorage wrapper
Article URL: https://github.com/brysonbw/ls-service Comments URL: https://news.ycombinator.com/item?id=50383396 Points: 1 # Comments: 0
37. Show HN: Umbel – Remote-control agent CLIs over tmux
Been using this in all my sessions the last months. Opus 5.5 (over MCP) spawns Deepseek flash agents - and codex for audits etc. Very curious to hear what the HN crowd thinks/makes of it. Comments URL: https://news.ycombinator.com/item?id=50383189 Points: 1 # Comments: 0
38. SRRM: A Retrieval-Reconstruction Benchmark for Evaluating Long-Context Memory in Small Language Models
arXiv:2611.05386v1 Announce Type: new Abstract: Small Language Models (SLMs) are increasingly deployed on consumer-grade hardware, yet their ability to preserve information across progressively longer contexts remains insufficiently understood. Existing benchmarks primarily evaluate retrieval-oriented tasks, providing limited insight into whether successful retrieval reflects comprehensive long-context memory preservation. To address this limitation, this paper introduces Simple Retrieval-Reconstruction Memory (SRRM), a lightweight and fully reproducible benchmark that jointly evaluates Targeted Retrieval and Global Information Reconstruction. This study also proposes the Memory Preservation Ratio (MPR), a benchmark-level metric that quantifies the relationship between retrieval and reconstruction performance. SRRM was evaluated using a synthetic dataset on three Qwen2.5 models (0.5B, 1.5B, and 3B parameters) across context loads ranging from 10 to 100 factual statements. Experimental results show that reconstruction performance deteriorated substantially faster than retrieval accuracy as contextual information increased, indicating that retrieval alone is insufficient to characterize long-context memory preservation. Among the evaluated models, Qwen2.5-1.5B achieved the highest retrieval accuracy (86.7%), reconstruction performance (45.7%), and MPR (0.53), outperforming the larger Qwen2.5-3B model (72.2%, 18.5%, and 0.26, respectively). These findings demonstrate that increasing parameter count does not necessarily improve long-context memory preservation. Overall, SRRM provides a reproducible and extensible benchmark for systematically evaluating long-context memory while revealing complementary memory behaviors that remain hidden under retrieval-only evaluation frameworks.
39. Stacking-Based Weighting for Large Language Bayes in M-Open Settings
arXiv:2611.03966v1 Announce Type: new Abstract: Large Language Bayes (LLB) transforms a natural language problem description into a collection of candidate probabilistic programs and performs inference within each model, combining their predictions using Bayesian model averaging. While this provides a flexible framework for probabilistic inference without manual model specification, the resulting candidate model set is often misspecified in practice, and likelihood-based aggregation can concentrate on a single model, under-representing uncertainty across the sampled models. We propose a stacking-based aggregation method for LLB that selects model weights by optimizing predictive performance rather than marginal likelihood, while preserving the original language-model-driven model generation and within-model inference pipeline. Because leave-one-out predictive quantities are intractable in latent-variable models, we derive a variational lower bound that yields a tractable ELBO-based stacking objective. Across real-world probabilistic inference tasks, our method achieves comparable predictive performance to Bayesian model averaging while yielding more dispersed model weights, reducing concentration on a single candidate model under misspecification.
40. The Wizard of Open Weights: Chris Alexiuk Makes AI Easy With Open Models
Chris Alexiuk has a simple rule about open source AI: If you believe in it, be loud about it. “We should be shouting about this technology from the rooftops,” said Alexiuk, a Toronto-based senior product research engineer in deep learning at NVIDIA who specializes in making open large language models (LLMs) practical for working developers. […]
41. Towards Efficient Whole Page Ranking with One-Pass Inference
arXiv:2611.05542v1 Announce Type: new Abstract: Whole-page ranking jointly optimizes advertisements and organic content to balance business outcomes and user engagement under complex layout and business constraints. Existing non-autoregressive (NAR) approaches enable efficient page generation but typically score candidate items independently of previously selected items, limiting their ability to capture order-dependent interactions. Autoregressive (AR) methods address this limitation through sequential conditioning, but incur substantial inference overhead that challenges large-scale deployment. Bridging this quality-efficiency gap remains a challenge in industrial whole-page ranking. We propose Multi-Item Scoring (MIS), a prefix-aware NAR-style framework that incorporates sequential context into efficient multi-item scoring. MIS integrates global candidate interactions, localized prefix conditioning, and position-aware scoring within a shared representation architecture. By computing reusable representations once per request and performing lightweight prefix-conditioned scoring during constrained beam search, MIS captures important sequential dependencies without repeated full-model inference. Experiments on KuaiRand-1K and RL4RS demonstrate that MIS approaches the generation quality of an autoregressive CGA baseline while achieving 4.07x and 3.54x faster CPU decoding, respectively. In a randomized evaluation on a large-scale social-video platform, MIS improves the realization metric by 0.104% and watch time by 0.058%, while reducing ranking-service CPU consumption by 9.73%. These results demonstrate that prefix-aware multi-item scoring can narrow the quality gap between NAR and AR generation while retaining the efficiency required for industrial scalability.
42. Towards Robust Long-Lived Agents: Benchmarking Experience Lifecycles in Evolving Environments
arXiv:2611.05450v1 Announce Type: new Abstract: As agents become more capable, they are moving beyond one-off tasks toward sustained assistance across successive user goals. Such assistance requires agents to discover how unfamiliar environments work, reuse what they learn across tasks, and keep that experience current as environments evolve. Existing benchmarks increasingly study experience reuse and adaptation, but provide limited insight into which parts of accumulated experience agents revise, retain, or lose. To address this gap, we introduce ReScopeBench, a benchmark for the experience lifecycle across 18 interactive environments. It distinguishes structural experience about how an environment is organized and accessed from rule experience about reusable principles and constraints that guide decisions across tasks. To study how experience is acquired, reused, and revised, we propose an experience-first construction approach. It builds task streams around shared experience and introduces silent environmental changes. We evaluate task performance and track changes in saved experience throughout each stream. Across 20 models, we find that retained experience improves task performance, but agents struggle to keep it current as environments change. Even when the required experience is correctly recorded, agents do not always complete tasks successfully. These findings identify selective maintenance and effective use as complementary challenges for long-lived agents.
43. University of Utah retains operational sovereignty as Redtail expands research access
Operational sovereignty gives universities authority over how shared computing resources are allocated, managed and secured. The University of Utah’s Redtail AI factory supercomputer brings advanced computing capabilities to a statewide community. Supporting researchers, students, administrators and external partners through that infrastructure requires more than computing capacity, according to Manish Parashar (pictured), chief AI officer of […] The post University of Utah retains operational sovereignty as Redtail expands research access appeared first on SiliconANGLE.
44. VirtualTrialHarness: Coordinating Agentic Operationalization of Virtual Trial Protocols Across Complex Clinical Schemas
arXiv:2611.05551v1 Announce Type: new Abstract: Virtual trials use routinely collected clinical data to evaluate treatment strategies, but translating a study protocol into an executable analysis requires determining how clinical concepts and study requirements should be represented in the available data and applied consistently across the workflow. These decisions depend on local data representations and operational conventions and can involve source selection, patient and encounter identity, linkage, and temporal interpretation. Here we introduce VirtualTrialHarness (VTHarness), an agent framework that coordinates virtual-trial operationalization by selecting reusable procedural skills for each workflow stage and uses explicit contracts to maintain consistent study decisions across stages. We also introduce VTBenchmark for evaluating source selection and cohort construction. Across three complex synthetic environments, VTHarness successfully produced all required cohort outputs matching the references in 95.1% of workflows after requirement clarification, compared with 60.4% using retrieval from the same procedural corpus. Procedural skills improved source recovery across three language models and increased source overlap with analyst code in 13 Mayo Clinic cases. Site-specific adaptation enabled reference-matched cohort execution across eight MIMIC-IV tasks. Within the Mayo Clinic Virtual Trials Program, the framework supported investigator review of cohort construction and endpoint analysis. These findings support coordinating procedural guidance with workflow decisions to improve virtual-trial implementation across complex clinical schemas.