# butterfly Published articles for butterfly. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## The (Butter)fly Dream: Linear Separability of World-Model Rollouts in a Connectome-Constrained Visual Model DevFeed: [The (Butter)fly Dream: Linear Separability of World-Model Rollouts in a Connectome-Constrained Visual Model]() Original publisher: [Read original article]() Author: Junichiro Niimi Published: 2026-11-03T05:00:00Z Content type: article Language: en Sources: [arXiv: Machine Learning]() Topics: [World models](), [Computer vision](), [model architecture](), [few-shot](), [machine learning overfitting]() Tags: [about](), [abstract](), [all](), [arxiv](), [brain](), [butterfly](), [can](), [column](), [cross](), [cross-validation](), [cs-lg](), [defects](), [detection](), [dream](), [dreams](), [fly](), [linear](), [model](), [published](), [q-bio-nc](), [world](), [world-model]() ### AI overview A connectome-constrained model of the fly visual system, trained only to estimate optic flow, can distinguish generated visual sequences from real footage using a linear classifier on its cell-type response statistics. Across 50 trained models, all nine continuations from three world models and three temperatures were separable from MPI-Sintel sequences (AUC 0.85-0.99), including unseen sequences; the same networks before training performed substantially worse. This separability did not track pixel error. In graded tests that shuffled the order of real frames, one adjacent swap among 14 generated frames yielded AUC 0.73 for trained responses, versus 0.54 before training and 0.56 for pixel statistics. A test over the first k generated frames helped order world models when full-sequence AUCs saturated. These findings are limited to the described model, task, footage, and probes. ### Source excerpt arXiv:2611.01866v1 Announce Type: new Abstract: Zhuangzi dreamt he was a butterfly and, on waking, could not tell which was the dream. We ask a version of his question of a model of the fly brain: placed inside a world generated by a world model, can a fixed brain tell that the world is a dream? We treat it as a question about what a connectome-constrained model trained on one task makes available beyond that task: do its responses make the defects of a dream linearly decodable? The observer is a published model of the fly visual system trained only for optic-flow estimation; reality is the MPI-Sintel movie on the model's 721-column retina; dreams are continuations of five real frames by three small world models (a persistence random walk, an MDN-RNN and an RSSM) at three temperatures. The judge is a classifier two-sample test on temporal statistics of cell-type responses, with sequence-level cross-validation. In the trained observer's responses all nine dreams are separable from reality (AUC 0.85-0.99, mean over 50 trained models), in those of the same network before training much less so (0.58-0.75 for the learned world models; paired difference 0.20-0.37), equally on sequences the observer never saw, and in an order unrelated to pixel error. Because these AUCs are near ceiling, we add graded probes made of each window's own real frames with only their order broken. Swapping one adjacent pair of the 14 generated frames gives 0.73 on the trained responses against 0.54 before training and 0.56 for pixel statistics; the trained curve reaches 0.75 at 1.1 swapped pairs of seven, the pixel statistics at 4.7, and the untrained network not at all. A detection horizon, the test on the first k generated frames, orders the world models where the full-length AUC saturates. ## SpectrumQuant: Spectrum Alignment for LLM Quantization DevFeed: [SpectrumQuant: Spectrum Alignment for LLM Quantization]() Original publisher: [Read original article]() Author: Dwaipayan Haldar, Indranil Patra, Azhar Yousuf, Chandra Sekhar Seelamantula Published: 2026-10-26T04:00:00Z Content type: article Language: en Sources: [arXiv: Machine Learning]() Topics: [quantization](), [Post-training optimization](), [Large Language Model]() Tags: [activation](), [alignment](), [butterfly](), [covariance](), [cs-lg](), [data](), [gpu](), [gpu-memory](), [large-language-models](), [large-language-models-llms](), [llama](), [llm](), [llm-quantization](), [llms](), [mistral](), [mixing](), [optimization](), [post-training](), [quantization](), [rotation](), [spectrum](), [stat-ml]() ### AI overview SpectrumQuant constructs a quantization-friendly orthogonal transform in closed form by estimating activation covariance from calibration data and aligning the spectrum. The paper frames useful rotations as alignment plus mixing, with Hadamard matrices as a special case and a class of butterfly matrices also fitting the framework. On Llama and Mistral models, the authors report competitiveness with state-of-the-art rotation strategies, using less than one-third the GPU memory and significantly less time than learned-rotation methods such as SpinQuant and DartQuant. ### Source excerpt arXiv:2610.20303v2 Announce Type: replace Abstract: Orthogonal transforms have emerged as an effective tool for low-bit quantization of large language models (LLMs), as they reshape activation distributions without changing the underlying model function. Existing approaches typically rely either on fixed data-agnostic transforms, such as Hadamard rotations, or on calibration-time optimization of learned rotations. We study rotation selection from first principles and show that maximizing covariance leads to spectral alignment and provides a principled criterion for constructing quantization-friendly orthogonal transforms. Based on this insight, we propose SpectrumQuant, a closed-form rotation method that estimates activation covariance from calibration data and uses it to construct a covariance-aware orthogonal transform for post-training quantization. Our formulation characterizes quantization-friendly rotations as compositions of alignment and mixing components, while recovering Hadamard matrices as a special case. Interestingly, a class of butterfly matrices belongs to this category. In the realm of LLM quantization, SpectrumQuant bridges the gap between data-agnostic structured rotations and learned rotations, providing an effective and efficient data-aware alternative for post-training quantization. Experiments on Llama and Mistral models show that SpectrumQuant is competitive with state-of-the-art rotation strategies while requiring less than one-third of the GPU memory and significantly less time than learned rotations approaches like SpinQuant and DartQuant. ## Stop Drawing Scientific Claims from LLM Social Simulations Without Robustness Audits DevFeed: [Stop Drawing Scientific Claims from LLM Social Simulations Without Robustness Audits]() Original publisher: [Read original article]() Author: Jinyi Ye, Lei Cao, Ding Chen, Emilio Ferrara Published: 2026-10-23T04:00:00Z Content type: opinion Language: en Sources: [arXiv: Artificial Intelligence]() Topics: [Simulation](), [Large Language Model](), [LLM security]() Tags: [abstract](), [agent](), [agents](), [also](), [artifacts](), [arxiv](), [audit](), [audits](), [based](), [butterfly](), [call](), [can](), [cascade](), [case-studies](), [cs-ai](), [cs-cy](), [cs-ma](), [drawing](), [llm](), [physics-soc-ph](), [robustness](), [simulations](), [social](), [stop]() ### AI overview The abstract reports that small design changes in LLM social simulations can substantially alter outcomes: persona format and game-instruction framing shifted cooperation by up to 76 percentage points in a repeated Prisoner's Dilemma, while network homophily and hub assignment consistently shifted polarization metrics in an echo-chamber simulation. Sensitivity varied by model; one perturbation caused a 76-point shift in one frontier model but only 1 point in another. The authors introduce TRAILS, a robustness-audit taxonomy covering agent, interaction, and system design, and argue that audits should be measured for each claim and model before simulations are used to explain mechanisms, evaluate interventions, or inform decisions. ### Source excerpt arXiv:2605.18890v2 Announce Type: replace-cross Abstract: The scientific claims drawn from LLM social simulations should be no stronger than the robustness audits that support them. Generative agents bring new expressive power to agent-based modeling, enabling simulations of collective social processes like cooperation, polarization, and norm formation. Yet they also introduce complexity through additional architectural choices, such as agent specification, memory representation, interaction protocols, and environment design. Small perturbations that appear minor to researchers can cascade into macro-level outcomes through repeated interaction, creating a "butterfly effect." Consequently, scientific claims drawn from LLM social simulations may reflect implementation artifacts rather than the social mechanisms being modeled. We support this position with two case studies: a repeated Prisoner's Dilemma and a social media echo chamber simulation. Across multiple models, minor perturbations in persona format and game-instruction framing shift cooperation rates by up to 76 percentage points, while network homophily and hub assignment produce significant and consistent shifts in polarization metrics. We also find that sensitivity is unevenly distributed across both architectural choices and model families: the same perturbation that produces the 76 pp shift in one frontier model only shifts another by 1 pp. Robustness is therefore a property that should be measured per claim and per model, not assumed. To address this validation gap, we introduce TRAILS (Taxonomy for Robustness Audits In LLM Simulations), a robustness-audit taxonomy spanning three levels of simulation design: agent (micro-level), interaction (meso-level), and system (macro-level). We call for robustness to become a first-order validation requirement before LLM social simulations are used to explain mechanisms, evaluate interventions, or inform decisions. ## Generative AI to quantify uncertainty in weather forecasting DevFeed: [Generative AI to quantify uncertainty in weather forecasting]() Original publisher: [Read original article]() Author: Google AI (noreply@blogger.com) Published: 2024-03-29T18:03:00Z Content type: article Language: en Sources: [Google Research]() Topics: [Climate / Weather / Ocean Modeling](), [Generative AI]() Tags: [ai](), [ai-model](), [ai-technology](), [butterfly](), [climate](), [diffusion](), [diffusion-models](), [extreme-weather](), [forecasting](), [generative](), [generative-ai](), [google](), [google-research](), [machine-learning](), [mit](), [physics](), [probabilistic](), [simulation](), [uncertainty](), [weather](), [weather-forecasting]() ### AI overview SEEDS uses a diffusion model to expand one or two physics-based weather forecasts into large ensembles at far lower computational cost than running a numerical weather model repeatedly. In the reported results, its ensembles matched or exceeded physics-based ensembles on several skill metrics and better represented rare tail events. At 2° resolution, it generated 256 members in three minutes on Google Cloud TPUv3-32 instances. Unlike a pointwise Gaussian model, SEEDS models the joint atmospheric distribution and captures spatial and cross-field correlations. In a seven-day forecast case for the 2022 European heat event, it generated 16,384 members from two seed forecasts and covered observed conditions that none of the operational ensemble's 31 members reached. These findings describe the study's reported cases; SEEDS is presented as a hybrid approach that supplements, rather than replaces, physics-based forecasts. ### Source excerpt Posted by Lizao (Larry) Li, Software Engineer, and Rob Carver, Research Scientist, Google Research Accurate weather forecasts can have a direct impact on people's lives, from helping make routine decisions, like what to pack for a day's activities, to informing urgent actions, for example, protecting people in the face of hazardous weather conditions. The importance of accurate and timely weather forecasts will only increase as the climate changes. Recognizing this, we at Google have been investing in weather and climate research to help ensure that the forecasting technology of tomorrow can meet the demand for reliable weather information. Some of our recent innovations include MetNet-3, Google's high-resolution forecasts up to 24-hours into the future, and GraphCast, a weather model that can predict weather up to 10 days ahead. Weather is inherently stochastic. To quantify the uncertainty, traditional methods rely on physics-based simulation to generate an ensemble of forecasts. However, it is computationally costly to generate a large ensemble so that rare and extreme weather events can be discerned and characterized accurately. With that in mind, we are excited to announce our latest innovation designed to accelerate progress in weather forecasting, Scalable Ensemble Envelope Diffusion Sampler (SEEDS), recently published in Science Advances. SEEDS is a generative AI model that can efficiently generate ensembles of weather forecasts at scale at a small fraction of the cost of traditional physics-based forecasting models. This technology opens up novel opportunities for weather and climate science, and it represents one of the first applications to weather and climate forecasting of probabilistic diffusion models, a generative AI technology behind recent advances in media generation. The need for probabilistic forecasts: the butterfly effect In December 1972, at the American Association for the Advancement of Science meeting in Washington, D.C., MIT meteorology pr