NormAct: A Benchmark for Hidden Social Norm Compliance in Embodied Planning
Shiyun Zhao, Xinwei Song, Tianyu Guo, Xiaomeng Gao, Mingyuan Liu
実装難易度
Hard
推論・学習コスト
High
想定用途
技術検証・論文読解補助
概要
Multimodal large language models (MLLMs) are increasingly deployed as embodied planners in egocentric environments, where task success requires not only achieving instructed goals but also acting in socially appropriate ways. While explicit goals may render certain actions optimal, implicit social norms often impose hidden constraints. Existing evaluations typically focus on explicit goal…
何が新しいか
Multimodal large language models (MLLMs) are increasingly deployed as embodied planners in egocentric environments, where task success requires not only achieving instructed goals but also acting in…
何に使えるか
技術検証・論文読解補助
実装情報
- Paper URL
- あり
実装チェックリスト
実装または配布ページ
要確認Paper onlyの可能性があるため再実装前提で確認してください。
一次情報リンク
OKPaper
検証しやすさ
要確認公式実装が見つからないため、論文から再実装する前提です。
計算資源
要確認学習や高解像度推論ではGPUメモリと実行時間に注意が必要です。
ライセンス
未取得配布元のLICENSE、モデルカード、Paperの利用条件を確認してください。
商用利用
未取得研究利用限定、データセット由来制限、API規約の有無を確認してください。
自社データで試すなら
製造業・材料開発のExcel/CSVデータに落とし込むための最初の手順です。
- 1まず自社データを、入力条件、目的変数、評価したい指標に分けて整理します。
- 2LightGBMやRandom Forestなどのベースラインを先に作り、この手法と比較します。
- 3評価指標はR2/RMSE、AUC、異常検知の再現率、実験回数削減率など、現場の意思決定に近いものを選びます。
- 4SHAPや特徴量重要度で、効いている因子が物理・化学・工程知識と矛盾しないか確認します。
実装難易度
Hard - 公式実装が見つからないため、論文から再実装する前提です。
必要リソース
- GPU目安: High
- データセット: 論文・リポジトリ側の指定を確認してください。
- 学習要否: 再学習や評価環境の準備が必要になる可能性があります。
- 学習や高解像度推論ではGPUメモリと実行時間に注意が必要です。
実務で使う場合の注意点
- ライセンスと商用利用条件は、Paper / GitHub / Hugging Face の配布元で確認してください。
- 精度、再現性、計算コストはデータセットや評価条件に依存します。
- 個人情報や機密データを扱う場合は、入力データの保存先と外部API利用条件を確認してください。
関連記事
RSICCLLM: A Multimodal Large Language Model for Remote Sensing Image Change Captioning
Remote Sensing Image Change Captioning (RSICC) aims to describe changes between bi-temporal remote sensing ima
DMV-Bench: Diagnosing Long-Horizon Multimodal Agents' Visual Memory with Incidental Cue Injection
Research on agent memory has matured rapidly, but almost entirely on the text side: few existing benchmarks as
ReasonCLIP-58M: Visually Grounded Commonsense Reasoning Supervision for CLIP
この研究では、視覚的共感推論をサポートする基盤を提案する。これは、CLIPなどのモデルを再トレーニングし、多様な推論タスクに適応化させることによって、視覚的に共感推論を実現する。
Beyond NL2Code: A Structured Survey of Multimodal Code Intelligence
While Large Language Models (LLMs) have substantially advanced text-to-code synthesis, many real programming t