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。
Memory capacity and bandwidth are major bottlenecks in LLM inference, motivating weight compression beyond conventional 4-bit quantization. Vector quantization (VQ) enables approximately 2-bit weight…
Computing molecular free-energy landscapes over collective variables (CVs) is central to understanding molecular mechanisms, but poses a difficult learning problem without training data available a…
In retinal vessel segmentation, plain Vision Transformers (ViTs) are commonly expected to perform worse than convolutional neural networks (CNNs) on thin vessels, for reasons such as their coarse…
Phase I dose-finding trials must learn from sparse toxicity observations while limiting exposure to overly toxic doses. We develop a single reinforcement-learning policy for trials with different…
Can a vision-language-action (VLA) policy signal likely task failure from its own action probabilities? We introduce CalVLA, which shapes native action-token confidence using successful and failed…
Before foraging, desert ants perform learning walks (LWs) to acquire visual memories around their nest. With experience, these walks expand, become straighter, and contain fewer visual scans.
Humans possess a remarkable ability to draw inferences from limited data. While some inferences are made intuitively from past experience, others require online reasoning based on novel information…
Large language model (LLM) agents increasingly rely on safeguards to prevent unauthorized actions while preserving legitimate service. However, existing risk assessments often focus on individual…
Production model development requires repeated architecture design, platform integration, debugging, training, and evaluation. We present an agent stack deployed at LinkedIn with two independently…
The hardest part of understanding a long video is not pinpointing any one moment but synthesizing many, e.g., answering a question that requires tracking what develops across hours of content and…
Future dynamic novel-view synthesis requires extrapolating multiple motion regimes within the same scene. A single dynamics model may not simultaneously represent stationary regions, coordinated…
Unlearning aims to make a large language model behave like one retrained without its forget set. Prevailing likelihood-suppression objectives, however, lack an explicit forgetting target: they…
Defense systems for maritime assets must distinguish approaching benign craft from uncrewed surface vessels (USVs) with sufficient warning to act. Shore cameras using computer vision can separate…
AI "humanizers" rewrite machine-generated text so that detectors classify it as human-written. Public leaderboards score evasion and embedding similarity, but not whether checkable details survive.
With model weights fixed, we measure how much of an entity-replacement response transfers between English factual questions, beyond whether answers change. We add the hidden-state change from two…
Multi-image vision-language models (VLMs) incur substantial computation when processing long visual sequences. Compressing these sequences must preserve complementary evidence across images and keep…
Self-verification and aggregation have emerged as popular approaches for bridging the gap between pass@1 and pass@N. However, increasing the number of candidates does not consistently narrow the gap…
AI systems embed majority-group defaults into training data, evaluation metrics, and category definitions, producing documented harms for queer communities including erasure, misclassification, and…
Retrieval-augmented generation (RAG) systems have seen growing adoption in embodied settings as language-queryable memories of the environment. However, existing approaches often assume a reliable…
Reinforcement learning (RL) policies often outperform classical controllers in simulation, yet rarely reach production in safety-critical processes. The barrier is that there is no principled way to…
Two Boolean functions are considered NPN equivalent if one can be transformed into the other by negating inputs, permuting inputs, or negating the output. Efficient NPN classification is valuable in…
Irregularly-spaced sequential data, such as transaction logs, encode multi-scale temporal structure that is critical for prediction but difficult to recover explicitly with standard sequence models,…
In this paper, we consider a class of possibly nonmonotone finite-sum composite inclusions in which evaluating all components or updating all coordinates at every iteration can be computationally…
Speech language models encode audio at a much higher rate than text, causing the KV cache to dominate the memory, bandwidth, and decoding cost of long-form generation. KV retrieval can reduce…