NormGym: Evaluating Normative Competence through Social Interaction
cs.AI
Submitted: 2026-11-05
Updated: 2026-11-05
Terminology
Sources
- Melting Pot 2.0
- Qwen3-VL Technical Report
- ARC-AGI-3: A New Challenge for Frontier Agentic Intelligence
- LoRA: Low-Rank Adaptation of Large Language Models
- Building Machines That Learn and Think Like People
- Ministral 3
- Learning and Sustaining Shared Normative Systems via Bayesian Rule Induction in Markov Games
- Corrupted by Reasoning: Reasoning Language Models Become Free-Riders in Public Goods Games
- NOPA: Neurally-guided Online Probabilistic Assistance for Building Socially Intelligent Home Assistants
- Emergence of Social Norms in Generative Agent Societies: Principles and Architecture
- EgoNormia: Benchmarking Physical Social Norm Understanding
- Maximum Likelihood Constraint Inference for Inverse Reinforcement Learning
- Agents of Chaos
- Human-Level Reinforcement Learning through Theory-Based Modeling, Exploration, and Planning
- Benchmarking World-Model Learning with Environment-Level Queries
- ReAct: Synergizing Reasoning and Acting in Language Models
- NormViz: A Benchmark and Framework for Grounding Multimodal Reasoning in Global Cultures
- NormAct: Benchmarking Embodied Agents' Proactive Compliance with Unspoken Social Norms
Related papers
- Multi-Objective Structured Pruning of LLMs for Latency and Model Size Optimization
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Embeddings for Preferences, Not Semantics
- Can We Trust Item Response Theory for AI Evaluation?
- Diagnosing Task Insensitivity in Language Agents