論文の概要: Generative AI collective behavior needs an interactionist paradigm
- arxiv url: http://arxiv.org/abs/2601.10567v1
- Date: Thu, 15 Jan 2026 16:29:23 GMT
- ステータス: 翻訳完了
- システム内更新日: 2026-01-16 19:43:19.227772
- Title: Generative AI collective behavior needs an interactionist paradigm
- Title(参考訳): ジェネレーティブAI集団行動は相互作用主義的パラダイムを必要とする
- Authors: Laura Ferrarotti, Gian Maria Campedelli, Roberto Dessì, Andrea Baronchelli, Giovanni Iacca, Kathleen M. Carley, Alex Pentland, Joel Z. Leibo, James Evans, Bruno Lepri,
- Abstract要約: 我々は,大規模言語モデル(LLM)に基づくエージェントの集団行動を理解することが,調査の不可欠な領域であると主張している。
我々は,LSMに基づく集団の開発と展開に不可欠な4つの方向性を提案し,議論する。
- 参考スコア(独自算出の注目度): 18.75814761446284
- License: http://creativecommons.org/licenses/by/4.0/
- Abstract: In this article, we argue that understanding the collective behavior of agents based on large language models (LLMs) is an essential area of inquiry, with important implications in terms of risks and benefits, impacting us as a society at many levels. We claim that the distinctive nature of LLMs--namely, their initialization with extensive pre-trained knowledge and implicit social priors, together with their capability of adaptation through in-context learning--motivates the need for an interactionist paradigm consisting of alternative theoretical foundations, methodologies, and analytical tools, in order to systematically examine how prior knowledge and embedded values interact with social context to shape emergent phenomena in multi-agent generative AI systems. We propose and discuss four directions that we consider crucial for the development and deployment of LLM-based collectives, focusing on theory, methods, and trans-disciplinary dialogue.
- Abstract(参考訳): 本稿では,大規模言語モデル(LLM)に基づくエージェントの集団行動を理解することが,リスクと利益の観点から重要な意味を持ち,多くのレベルで社会として私たちに影響を与えることを論じる。
従来の知識と組込み値が社会的文脈とどのように相互作用し、マルチエージェント生成型AIシステムにおいて創発的現象を形成するかを体系的に検討するために、LLMの独特な性質、すなわち、広範囲な事前学習された知識と暗黙的な社会的先行性による初期化と、文脈内学習による適応能力は、代替理論の基礎、方法論、分析ツールからなる相互作用論パラダイムの必要性を動機付けている、と我々は主張する。
我々は,LLMに基づく集団の開発と展開に不可欠な4つの方向性を,理論,手法,学際的対話に焦点をあてて提案し,議論する。
関連論文リスト
- Understanding LLM Agent Behaviours via Game Theory: Strategy Recognition, Biases and Multi-Agent Dynamics [1.6487772637295166]
我々はFAIRGAMEフレームワークを拡張し、繰り返しの社会的ジレンマにおける大規模言語モデル(LLM)の振る舞いを評価する。
LLMは、言語フレーミングが建築的差異に匹敵する効果を発揮することを示し、体系的、モデル的、言語に依存した行動意図を示す。
論文 参考訳(メタデータ) (2025-12-08T11:40:03Z) - Simulating Students with Large Language Models: A Review of Architecture, Mechanisms, and Role Modelling in Education with Generative AI [0.8703455323398351]
大規模言語モデル(LLM)を用いた教育環境における学生の行動のシミュレートに関する研究のレビュー
LLMをベースとしたエージェントが学習者の古型をエミュレートし、指導的な入力に反応し、マルチエージェントの教室シナリオ内で相互作用する能力に関する現在の証拠を示す。
本稿では,カリキュラム開発,教育評価,教員養成におけるシステムの役割について検討する。
論文 参考訳(メタデータ) (2025-11-08T17:23:13Z) - A Survey of Frontiers in LLM Reasoning: Inference Scaling, Learning to Reason, and Agentic Systems [93.8285345915925]
推論(Reasoning)は、論理的推論、問題解決、意思決定を可能にする基本的な認知プロセスである。
大規模言語モデル(LLM)の急速な進歩により、推論は高度なAIシステムを区別する重要な能力として浮上した。
我々は,(1)推論が達成される段階を定義するレジーム,(2)推論プロセスに関与するコンポーネントを決定するアーキテクチャの2つの側面に沿って既存の手法を分類する。
論文 参考訳(メタデータ) (2025-04-12T01:27:49Z) - PersLLM: A Personified Training Approach for Large Language Models [66.16513246245401]
データ構築とモデルチューニングを改善するためのフレームワークPersLLMを提案する。
データ利用が不十分な場合には、Chain-of-Thoughtプロンプトやアンチインダクションといった戦略を取り入れます。
厳密な振舞いパターンを設計し,モデルの性格の特異性とダイナミズムを高めるために自動DPOを導入する。
論文 参考訳(メタデータ) (2024-07-17T08:13:22Z) - Transforming Competition into Collaboration: The Revolutionary Role of Multi-Agent Systems and Language Models in Modern Organizations [0.0]
本稿では,マルチエージェントシステム理論(SMA)と大規模言語モデル(LLM)に基づく計算エンティティがユーザインタラクションに与える影響について考察する。
提案手法では,大規模言語モデル (LLM) から発達したエージェントを用いて,行動要素を考慮したプロトタイピングを行う。
我々は,多エージェントシステム理論(SMA)と大規模言語モデル(LLM)に基づく革新的な利用に基づいて,組織戦略に有用なエージェントの開発の可能性を示す。
論文 参考訳(メタデータ) (2024-03-12T15:56:10Z)
関連論文リストは本サイト内にある論文のタイトル・アブストラクトから自動的に作成しています。
指定された論文の情報です。
本サイトの運営者は本サイト(すべての情報・翻訳含む)の品質を保証せず、本サイト(すべての情報・翻訳含む)を使用して発生したあらゆる結果について一切の責任を負いません。