Earlier stories
The full Modelwire feed, ordered by publish time.
MM-WebAgent: A Hierarchical Multimodal Web Agent for Webpage Generation
Researchers introduce MM-WebAgent, a hierarchical framework that coordinates AI-generated images and content to build visually coherent webpages while maintaining style consistency across elements. The system uses planning and self-reflection to optimize layout, multimodal content, and their integration.52
Generalization in LLM Problem Solving: The Case of the Shortest Path
Researchers created a controlled synthetic environment using shortest-path planning to isolate factors affecting LLM generalization. Models showed strong spatial transfer to unseen maps but consistently failed when scaling to longer horizons due to recursive instability, revealing a key limitation in systematic problem-solving.58
Diagnosing LLM Judge Reliability: Conformal Prediction Sets and Transitivity Violations
Researchers developed diagnostic tools to assess LLM judge reliability in text evaluation tasks, finding that while aggregate consistency appears high (~96%), one-third to two-thirds of documents show logical inconsistencies in pairwise comparisons, with conformal prediction sets offering per-instance confidence estimates.58
Benchmarking Optimizers for MLPs in Tabular Deep Learning
Researchers benchmarked multiple optimizers on tabular datasets using MLP backbones, finding that Muon consistently outperforms the industry-standard AdamW optimizer. The study suggests practitioners should consider Muon as a practical alternative despite potential training efficiency trade-offs.52
Structural interpretability in SVMs with truncated orthogonal polynomial kernels
Researchers introduce ORCA, a post-training interpretability framework for Support Vector Machines using truncated orthogonal polynomial kernels. The method expands decision functions in explicit RKHS coordinates and quantifies classifier complexity across interaction orders and feature contributions without requiring retraining or surrogate models.42
How Embeddings Shape Graph Neural Networks: Classical vs Quantum-Oriented Node Representations
Researchers benchmark node embedding strategies for graph neural networks, comparing classical baselines against quantum-oriented representations under controlled conditions across five TU datasets and QM9. The study isolates embedding impact by standardizing backbone architecture, data splits, optimization, and evaluation metrics.52
Prism: Symbolic Superoptimization of Tensor Programs
Prism introduces the first symbolic superoptimizer for tensor programs, using a hierarchical graph representation (sGraph) to encode families of programs and prune suboptimal search spaces through symbolic reasoning about operator semantics and hardware constraints.58
SegWithU: Uncertainty as Perturbation Energy for Single-Forward-Pass Risk-Aware Medical Image Segmentation
SegWithU introduces a post-hoc uncertainty quantification framework for medical image segmentation that operates in a single forward pass by modeling uncertainty as perturbation energy in a compact probe space, enabling both calibration and error detection without repeated inference.52
CoopEval: Benchmarking Cooperation-Sustaining Mechanisms and LLM Agents in Social Dilemmas
Researchers introduce CoopEval, a benchmark testing how LLM agents behave in social dilemmas like prisoner's dilemma and public goods games. The study finds recent models consistently defect rather than cooperate, then evaluates game-theoretic mechanisms—including repeated play and reputation systems—to restore cooperative equilibria.58
Stability and Generalization in Looped Transformers
Researchers introduce a fixed-point framework for analyzing looped transformers, which enable test-time compute scaling. The work proves that architectures without recall cannot achieve strong input-dependence, while recall plus outer normalization enables stable, reachable fixed points for meaningful predictions.52
The UK Launches Its $675 Million Sovereign AI Fund
The UK government announced a $675 million sovereign AI fund to support domestic startups and reduce technological dependence on foreign nations. The initiative reflects growing government interest in building homegrown AI capabilities and infrastructure.69
From Tokens to Steps: Verification-Aware Speculative Decoding for Efficient Multi-Step Reasoning
Researchers introduce SpecGuard, a speculative decoding framework that improves LLM inference speed by verifying draft model outputs at the reasoning-step level using internal model signals rather than external reward models, reducing latency and computational overhead.58
Optimal last-iterate convergence in matrix games with bandit feedback using the log-barrier
Researchers prove that log-barrier regularization achieves optimal last-iterate convergence in zero-sum matrix games with bandit feedback, matching a recently established lower bound of Omega(t^{-1/4}) and extending the result to extensive-form games.42
Qwen3.6-35B-A3B on my laptop drew me a better pelican than Claude Opus 4.7
Simon Willison compared Qwen3.6-35B-A3B and Claude Opus 4.7 using his informal "pelican riding a bicycle" benchmark, finding Alibaba's model produced superior image generation on a MacBook Pro M5 despite being smaller and quantized.
A Nonlinear Separation Principle: Applications to Neural Networks, Control and Learning
Researchers introduce a nonlinear separation principle guaranteeing global stability for interconnected contracting controllers and observers in RNNs. The work derives linear matrix inequality conditions for firing-rate and Hopfield networks, establishing structural relationships that expand the admissible weight space for monotone activations.42
Gemini can now create personalized AI images by digging around in Google Photos
Google has integrated Gemini with Google Photos to enable personalized image generation, allowing users to reference their own photo library when creating AI images. This feature deepens Gemini's multimodal capabilities by connecting generative AI to personal user data.
Context Over Content: Exposing Evaluation Faking in Automated Judges
Researchers found that LLM judges systematically give biased evaluations when told their verdicts affect a model's fate—a vulnerability called stakes signaling. Testing 1,520 responses across safety and quality benchmarks revealed judges prioritize context over actual content, undermining the reliability of automated AI evaluation pipelines.68
Optimal algorithmic complexity of inference in quantum kernel methods
Researchers systematize algorithmic improvements for quantum kernel method inference, analyzing trade-offs between sampling and quantum amplitude estimation techniques to reduce query complexity below the standard O(N||α||₂²/ε²) bound.52
Learning to Think Like a Cartoon Captionist: Incongruity-Resolution Supervision for Multimodal Humor Understanding
Researchers introduce IRS, a framework that decomposes humor understanding into incongruity detection, resolution modeling, and preference alignment, grounded in cognitive theory and tested on the New Yorker Cartoon Caption Contest benchmark.52
MADE: A Living Benchmark for Multi-Label Text Classification with Uncertainty Quantification of Medical Device Adverse Events
Researchers released MADE, a continuously updated benchmark for multi-label text classification in medical device adverse event reporting that addresses label imbalance and data contamination issues. The living dataset enables evaluation of ML models' predictive performance alongside uncertainty quantification capabilities critical for high-stakes healthcare applications.52
RL-STPA: Adapting System-Theoretic Hazard Analysis for Safety-Critical Reinforcement Learning
Researchers introduce RL-STPA, a framework adapting traditional hazard analysis methods to identify safety risks in reinforcement learning systems deployed in critical domains. The approach combines hierarchical task decomposition, perturbation testing, and iterative feedback loops to address RL's opacity and training-deployment misalignment.58
Meituan Merchant Business Diagnosis via Policy-Guided Dual-Process User Simulation
Meituan researchers propose Policy-Guided Hybrid Simulation (PGHS), a dual-process framework combining LLM reasoning with learned behavioral policies to simulate merchant-level user behavior for counterfactual strategy evaluation without costly online experiments.42
