MLINFO COLLECTION
日次更新の新着・公開記事を新しい順に掲載しています。
Retrieval-augmented generation (RAG) helps large language models (LLMs) answer factual questions beyond their training data, but it adds storage and inference costs. Distilling previously retrieved…
SpBench is an agent-native framework for cross-engine lakehouse benchmarking, in production use for benchmarking Snowpark, Snowpark Connect for Apache Spark, and other execution environments across…
Learning the continuous-time evolution of nonlinear dynamical systems relies on trajectory-based objectives, which require simulating the learned dynamics during training and backpropagating through…
LLM-as-a-judge scores are often distributionally shifted from human ratings on any given evaluation task, and post-hoc calibration (mapping LLM to human scores in a small labeled anchor set) has…
Analog computers solve problems through physical representations of mathematical variables and relations, often using the dynamics of electronic circuits and their associated signals. Many analog…
High-resolution video diffusion relies on attention over long token sequences, making attention efficiency critical to generation latency. Our analysis identifies repeated maximum reductions and…
When a multi-turn agent delivers a tool's structured records into the model's context and re-emits them each turn, every turn can corrupt data the tool computed correctly. We measure this in a…
As AI systems are increasingly deployed in high-stakes domains such as healthcare, effectively integrating human expertise with AI outputs is critical for reliable decision-making. A substantial body…
Performativity arises when prediction-informed decisions change the data-generating process, creating feedback between deployment and learning. This phenomenon occurs in policy-making and in…
Urban flood resilience emerges from complex interactions among natural, built, and socio-economic systems, yet most recovery policies implicitly assume that resilience mechanisms are spatially…
We study the online fair division of indivisible chores among $n$ agents under upgrade recourse, where earlier assignments may be revised but each chore previously assigned to an agent must be…
Amodal video object segmentation is fundamentally limited by the absence of datasets that combine real-world complexity with precise ground-truth annotations. To address this, we present Real Video…
A spectrum can be given to a model as a table in nanometers or as a plot in photon energy, and the decision should not depend on which one it sees. When the data needed for an answer are missing, the…
Joint-embedding predictive architectures (JEPAs) have recently emerged as general-purpose visual world models for control. A prominent example is LeWorldModel (LeWM), which trains end-to-end from…
Dense action detection requires precisely localising multiple overlapping actions in long untrimmed videos. Transformer-based methods achieve strong accuracy but their $O(T^2)$ self-attention cost…
We study how the layers in deep neural networks (NNs) converge, using Wilsonian Renormalization-Group (RG) ideas to organize the spectral picture. During training, these layers often develop…
Mixture-of-experts (MoE) models expand model capacity by activating only a subset of experts for each token. On accelerators with static execution plans, expert capacities and placement are fixed…
Attribution graphs, an emerging tool in mechanistic interpretability, use transcoders to decompose language model computations into sparse interpretable features connected by causal edges. However,…
We study the local differential geometry of a Fourier-based modular-addition circuit in a one-layer transformer whose algorithm has previously been reverse-engineered. For each of the five key…
AI red-teaming, adversarially probing generative AI systems to surface harms, has become a central evaluation method. However, prior work has highlighted that red-teaming is a fairly open-ended task,…
Smartphone applications increasingly require large language models to process long contexts, whereas high prefill latency and memory demands constrain local inference. Block-sparse attention reduces…
An agent's visible score can exceed its real progress because it gamed the score or because it simply made mistakes, so a useful reward-hacking monitor must tell the two apart. We built a sandboxed…
Tracking points that leave the camera's field of view (FoV) requires reasoning about unobserved scene dynamics, a fundamentally different problem from tracking visible correspondences. We present…
Data flywheels promise continuous improvement via insights from deployment data. Understanding how the deployed model succeeds and fails can steer subsequent data collection towards patching failures…