MLINFO COLLECTION
検索拡張生成、文書検索、社内ナレッジQAに関係する論文・実装をまとめます。
Numbers computed from trained models, such as an example's difficulty or a model's benchmark lead, change with seed and prompt. We ask when such measurements can be corrected or compared.
We describe the QMUL_AK submission to AISA-ArabicFC 2026. The system adapts the Arabic-English ALLaM-7B-Instruct model to function calling with low-rank adaptation (LoRA).
Latent multi-agent systems reduce inference cost by keeping intermediate reasoning and inter-agent communication in continuous representations.\ Reusable skills, however, are commonly supplied as…
Large language model agents are increasingly used for scientific discovery. However, existing benchmarks mix law inference with programming, score data fitting instead of prediction on unseen inputs,…
We propose a multimodal transformer framework for efficiently modeling long1 financial documents that combine structured tabular data with noisy, unstructured2 textual disclosures. The framework…
Large language models (LLMs) are widely used as automated judges, yet most alignment methods target aggregated human judgments and overlook variation among individual annotators. Existing approaches…
Retrieval-augmented generation (RAG) lets a large language model (LLM) answer questions by retrieving relevant passages from a large corpus. However, RAG struggles with multi-hop questions, which…
Large language models (LLMs) have shown increasing promise in research-level problem solving, yet a comprehensive collection of such problems is still missing, and evaluating them on research-level…
Large Language Models (LLMs) are increasingly used in multi-turn agentic workloads that require sustained interaction with complex environments. Existing reflective prompt optimizers such as GEPA…
Self-attention plays a key role in long-sequence modeling, but its quadratic time complexity fundamentally limits the scaling of Large Language Models. Recently, Multi-State Linear Attention (MSLA)…
In smart factory environments, passive entities such as humans and movable obstacles may interfere with autonomous robot operations, posing significant safety risks. Detecting potential conflicts…
Multi-robot control requires integrating distributed camera views while predicting separate action sequences. Most vision-language-action and world-action models target one robot.
As egocentric video becomes increasingly available for embodied learning, an important research question is how to turn this low-cost, scalable source of data into robot manipulation capabilities.…
Large language models (LLMs) are increasingly used to make and execute decisions that affect living systems. Yet, to our knowledge, ecological reasoning is not a focus of pre- or post-training.
Fine-tuning Large Language Models (LLMs) on instruction data can quietly erode their safety, even when no training example looks harmful. Because this safety-degrading data looks benign, standard…
Long-term memory enables Large Language Models (LLMs) to maintain coherent and personalized interactions across extended conversations. However, existing memory systems face two key challenges: (1)…
Language model performance depends strongly on the mixture of domains in the pretraining data corpus. Popular data mixing methods select these domain weights by training many small proxy models on…
Large language model (LLM)-driven agent scaffold evolution improves agent workflows through iterative candidate generation and evaluation. Evaluation both guides subsequent search and determines the…
Proprietary language models and their reasoning traces are vulnerable to API-based distillation, where an adversary trains a student model on the teacher model's outputs. To obtain evidence of…
Vision-Language-Action (VLA) policies demonstrate remarkable semantic generalization, yet remain highly vulnerable to out-of-distribution (OOD) visual shifts during execution. While existing…
We trained four widely used imitation-learning policies, ACT, Diffusion Policy, SmolVLA and $π_{0.5}$, on nested subsets of 12, 25, 50, 100 and 200 teleoperated demonstrations of one pick-and-place…
Variational inference with normalizing flows provides a flexible posterior approximation for Bayesian model calibration, but struggles when the posterior has several modes separated by regions of low…
Coverage-evaluation benchmarks quantify whether retrieved evidence covers the different pieces of information needed to satisfy an information need. Existing benchmarks mostly consider factual claims…
Knowledge distillation compresses large language models into compact students that preserve aggregate metrics, yet the mechanistic consequences of depth-reduction distillation remain largely…