学習・ファインチューニング
A
85 件中 31〜60 件目を表示
-
Beyond In-Distribution Metrics: A Systematic Out-of-Distribution Evaluation of Congenital Heart Disease Segmentation
-
Audio-Visual Turn-taking Prediction in Cocktail Party Scenarios
-
CLARE: Scalable Class-Incremental Continual Learning via a Sparsity-Based Framework
-
Beyond Measurement Metrics: A Human-Centered Framework for Semantic Validation of Network Traffic Classification
-
Can LLMs Follow the Pulse of a Crisis? Evaluating Crisis Sentiment in Bangladesh's July Uprising
-
MedPCFM-TED: One-Step Point Cloud Flow Matching for Implant Generation via Teacher-Guided Endpoint Distillation
-
TAME: Token Attribution and Masking for Emergent misalignment
-
LSREP: A Longitudinal State-Replay Protocol for Evaluating Conversational Memory, with ICE v2 as an Audited Local-First Architecture
-
Disentangling Representation Evolution in Transformers through Directional Decomposition
-
Inoculation Midtraining with Learned Neologisms
-
LLM-Based Schema-Aware Split Learning for Privacy-Preserving Mental Distress Prediction Across Heterogeneous Surveys
-
LongAgent: History-Guided Agentic Search for Longitudinal Outcome Prediction
-
K-Bench: a clinically calibrated benchmark for evaluating large language models in high-risk mental health conversations
-
AlgoEvo: Self-Evolving Agentic Search for Automated Algorithm Discovery
-
Navigating Sparse Evidence: Agentic Visual RAG via Explicit Context Selection and Consolidation
-
When the World Lies: Backdoor Attacks on Latent World Models for Downstream Control
-
AnthropicのMCP共同作者が来日、基調講演で語るこれからのMCP注力領域。AGNTCon＋MCPCon Japan 2026AnthropicのMCP共同作者が来日、今後1年はエージェント型メッセージング・Skills・認可に注力と講演Linux Foundation主催「AGNTCon＋MCPCon Japan 2026」(9月10日、東京渋谷) の基調講演に、MCP共同作者であるAnthropicのDavid Soria Parra氏が登壇。Claude上のMCP経由ツール呼び出しは累計約10億件、SDKダウンロードは月間5億件超と急成長を示し、今年MCPを完全ステートレス化したと説明した。今後12カ月は「MCP Tasks」による長期処理向けのエージェント型メッセージング、スキルをMCPサーバ経由で提供する「Skills over MCP」、エージェントの認可とアイデンティティの3領域に注力すると語った。
-
OpenAIやAnthropicなどAIベンダごとのAPIの違いを吸収し統合する「Agent Router」、Linux Foundation傘下で業界標準へEnvoy AI GatewayがAAIFに加盟し「Agent Router」に改名、LLM API差異を吸収する業界標準へLinux Foundation傘下のAgentic AI Foundation (AAIF) は、オープンソースの「Envoy AI Gateway」がAAIFに加盟し「Agent Router」に改名したと、9月10日のAGNTCon＋MCPCon Japan 2026基調講演で発表した。OpenAI、Anthropic、Gemini、Bedrock、Azure OpenAI など多数のAIサービスをOpenAI互換の単一APIで扱え、複数MCPサーバを束ねるMCP Gateway、フォールバックやトークン量制限、メトリクスに基づくルーティング、可観測性も備える。既にv1.1で本番利用可能で、AAIFは事実上の業界標準化を目指す。
-
Sequential Adapter Stacking for Cross-Lingual Low-Resource ASR
-
Merging the Knowledge of LLMs for Automatic Speech Recognition
-
A Language-Guided Multimodal Foundation Model for Zero-Shot and Multi-Task Brain Signal Analysis
-
Empathy Is Steerable but Multi-Axial: Mechanism Geometry and Persona Effects in LLMs
-
Principal-timestep Restricted Init via Sparse Matrix-decomposition in Flow-matching
-
Beyond AI Literacy: A Structured Review and Exploratory Meta-Analysis of Measures for Competent Generative-AI Use
-
IROH: Insightful Ranking Of Humor using Multi-Stage Hybrid Retrieval with Rationale-Distilled LLM Judges for JOKER 2026 Track Task 1 English
-
Multi-View Molecular Representation Learning with Hierarchical Graphs and Contextualized Fingerprints
-
Through the Eyes of the Beholder: Biometric and Demographic Conditioning for Multimodal Sexism Detection
-
Specifying Reward Functions for RL Without Environment Sampling
-
Option-Aware Retrieval and Task-Specific VLM Adaptation for Medical VQA
-
To Each Language Its Tokenizer: Modular Tokenizers for Efficient Multilingual LLMs