MLINFO COLLECTION
Vietnamese pre-trained encoders have improved rapidly, but the strongest ones are large, multilingual-vocabulary models or depend on external word segmentation, which makes them costly to deploy. We…
Uniform diffusion language models (UDLMs) generate text by iteratively denoising sequences of ordinary vocabulary tokens, any of which may be corrupted and subsequently revised. During training,…
Humanoid loco-manipulation requires an effective interface between high-level task policies and whole-body control. Existing interfaces typically rely on a single action representation for both task…
Harness optimization, which improves an agent's prompts, tools, and workflows, is an increasingly effective way to enhance agents built on frozen or closed-source LLMs. Leading approaches such as the…
Language 3D Gaussian Splatting enables open-vocabulary scene understanding by associating Gaussian primitives with language features. These semantic attributes increase model storage, while most…
Training generative models on synthetic data causes loss of diversity and a decrease in performance, a phenomenon known as model collapse. We study this degeneration spectrally in a controlled Markov…
AI-generated text detectors are increasingly deployed in consequential settings---peer review, education, admissions, journalism, and content moderation---where their outputs directly affect…
Natural-language software requirements often omit assumptions about performance, operating conditions, user behaviour, and domain constraints, affecting architecture, testing, and deployment. If…
Transformer-based models achieve strong performance on blind compressed image restoration (CIR) under unknown compression settings, but their densely stacked projection operators make deployment…
This paper describes our submission to the WMT 2026 Multilingual Instruction Shared Task. We adapt Qwen3-8B with LoRA using organiser-provided data, EuroAlpaca, a multilingual instruction dataset…
Thermal cameras can watch over older people living alone without capturing faces, and multimodal large language models (LLMs) can judge a short video without task-specific training. We evaluate a…
An $n$-player game in normal form is said to be of format $\boldsymbol{d} = (d_1, \dots, d_n)$ if the $i$-th player has $d_i$ options for $1 \le i \le n$, and is said to be balanced if $d_n - 1 \le…
Learning across medical imaging modalities requires making use of images without paired text and accommodating new domains while retaining existing representations. We propose a medical…
Reliable uncertainty estimation is important across visual recognition tasks, providing information about the confidence and reliability of model predictions. Many uncertainty estimation methods…
How do diffusion transformers (DiTs) represent image content and learn rules for reusing it to generalize? We develop an aggregator--predictor view of DiTs as input-adaptive local denoisers, in which…
Autonomous mobile robots operating in dense crowds and cluttered sites require $360^\circ$ perception to react to surrounding hazards. While panoramic LiDAR offers full-surround coverage, learning…
Literature-searching language agents are becoming a standard component of scientific assistants for the life sciences, yet there is little controlled evidence on which parts of their harness matter.…
In likelihood-free hypothesis testing (LFHT), given samples from two unknown distributions $P,Q$ on $[n]$ with $\mathrm{d_{TV}}(P,Q)\geq\varepsilon$, and from a third distribution $D\in\{P,Q\}$, the…
Object detection models rely on large-scale annotated datasets that require substantial manual annotation effort. The labeling costs further increase when annotation uncertainty is captured through…
Asked what to study, which career to pursue, or what product to build, large language models (LLMs) often recommend AI-related options. We ask where this tendency comes from by comparing models…
The CLARKE building and road damage assessment AI classifier was incorporated into the Catastrophic Unmanned Remote Sensing Exercise (CURSE) 2026, which trained 23 small uncrewed aerial systems…
Whole slide image (WSI) analysis enables computational diagnosis and prognosis, but gigapixel images and limited annotations make slide-level learning challenging. Multiple instance learning…
Compact chest X-ray (CXR) models that localize findings suit deployment, but box annotations are limited and report-aware multimodal models are expensive at inference. We introduce APED-CXR, which…
Understanding failures of autonomous systems is challenging, particularly when their behavior depends on black-box components such as neural networks. In this work, we make a distinction between two…