学習・ファインチューニング
A
118 件中 31〜60 件目を表示
-
VeriCUA: Verification as Agentic Exploration for Computer-Using Agents
-
Training Foundation Models with a Single Algorithm
-
Advancing Clinical Genomic Analysis with a Skill-Centred Agent
-
Algebraic World Model: From Exploration to Proofs
-
Glance: The Generator Is Its Own Reranker
-
Building a multiplayer game world with autonomous Fable-class modelsFable級の自律モデルで多人数参加型ゲーム世界を構築Simon Willisonが紹介する、Jacob SansburyによるSpawn構築の解説。釣りのような見出しだが、フロンティアモデルを用いたエージェント的エンジニアリングの最先端を探る興味深い内容だという。
-
SCARL: Sequence Confidence-Aligned Reinforcement Learning
-
Continuous-Action World Models: Learning Cross-Embodiment Spatiotemporal Dynamics Priors
-
Scaling Mathematical Proof Supervision from Structured Textbooks
-
Composing and Steering Diffusion Models with Generalized Feynman-Kac Correctors
-
Your LR Schedule May Not Be Principled: Effective Learning Rate Scheduling in Weight-Normalized Training
-
Latent Path Exploration via Potential-Guided Subtrajectory Balance
-
Latent Prediction Yields Cross-Subject Alignable Representations in Sensory-Evoked fMRI
-
Neighborhood Elites: Quality-Diversity Search for Uncertainty Quantification in Question Answering
-
Statistical Analysis of Synthetic Data with Verification: A Linear Regression Framework
-
Upcycling Language Models for Multiple Specialized Domains
-
Rethinking Pre-pretraining
-
Can LLMs Ignore What They Know? Measuring Knowledge Leakage in Persona-Based User Simulation
-
Surprise Alone Learns Lies: Evidence-Gated Continual Learning for LLMs from Streaming Claims
-
Sparse Fine-Tuning for Parameter-Efficient Adversarial Training
-
ARBITRA: Robust Document-Level ATT&CK Extraction via Multi-Expert Argumentation
-
FDSML: Formula-Driven Supervised Metric Learning with Parameter-Aware Triplet Loss
-
Safe Multi-Objective Reinforcement Learning: Zero Constraint Violation and Primal-Averaging Convergence
-
TaxoMFM: A Taxonomy-Aware Foundation Model for the Human Gut Microbiome
-
Multi-Solution Inverse Design of Pixelated Patch Antennas Using an Autoregressive Transformer
-
XLaw-JP-27B: A Lightweight Open Model Specialized for Japanese Law
-
Where Position Matters: Profiling-Guided Layer-Wise RoPE Allocation in Transformers
-
Event2Reward: Learning Robotic Rewards from Progress Relations
-
Uncertainty-Guided Cross-Modal Alignment for Diffusion-Based IoT Intrusion Detection
-
PALATE: Pretrained Autoregressive LATent Embeddings