Sensi: Learn One Thing at a Time -- Curriculum-Based Test-Time Learning for LLM Game Agents

• URL: http://arxiv.org/abs/2603.17683v1
• Date: Wed, 18 Mar 2026 12:59:26 GMT
• Title: Sensi: Learn One Thing at a Time -- Curriculum-Based Test-Time Learning for LLM Game Agents
• Abstract summary: We present Sensi, an agent architecture for ARC-AGI-3 game-playing challenge that introduces structured test-time learning.<n>We report results across two iterations: Sensi v1 solves 2 game levels using the two-player architecture alone, while Sensi v2 adds curriculum learning and solves 0 levels.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language model (LLM) agents deployed in unknown environments must learn task structure at test time, but current approaches require thousands of interactions to form useful hypotheses. We present Sensi, an LLM agent architecture for the ARC-AGI-3 game-playing challenge that introduces structured test-time learning through three mechanisms: (1) a two-player architecture separating perception from action, (2) a curriculum-based learning system managed by an external state machine, and (3) a database-as-control-plane that makes the agents context window programmatically steerable. We further introduce an LLM-as-judge component with dynamically generated evaluation rubrics to determine when the agent has learned enough about one topic to advance to the next. We report results across two iterations: Sensi v1 solves 2 game levels using the two-player architecture alone, while Sensi v2 adds curriculum learning and solves 0 levels - but completes its entire learning curriculum in approximately 32 action attempts, achieving 50-94x greater sample efficiency than comparable systems that require 1600-3000 attempts. We precisely diagnose the failure mode as a self-consistent hallucination cascade originating in the perception layer, demonstrating that the architectural bottleneck has shifted from learning efficiency to perceptual grounding - a more tractable problem.

Related papers

• EduClaw-Bench: A Long-Horizon Benchmark for Pedagogical LLM Agents with Simulated Learners [1.6405027086146922]
  tutoring is long-horizon, since a learner improves over days and weeks rather than in a single turn.<n>We introduce EduClaw-Bench, a benchmark that places an agent tutor in a continuous 30-day relationship with a simulated learner.
  arXiv  Detail & Related papers  (2026-08-04T06:47:26Z)
• EEVEE: Towards Test-time Prompt Learning in the Real World for Self-Improving Agents [64.96332056338923]
  EEVEE is the first multi-dataset test-time prompt learning framework for LLM agents.<n>It enables test-time prompt learning under real-world task streams.<n>EEVEE improves average multi-benchmark scores by 10.38 and 24.32 points over Qwen3-4B-Instruct and DeepSeek-V3.2.
  arXiv  Detail & Related papers  (2026-06-09T17:57:16Z)
• Autonomous Quantum Simulation through Large Language Model Agents [0.29165586612027233]
  Large language model (LLM) agents can autonomously perform tensor network simulations of quantum many-body systems.<n>We create autonomous AI agents that can be trained in specialized computational domains within minutes.
  arXiv  Detail & Related papers  (2026-01-15T08:50:57Z)
• Actial: Activate Spatial Reasoning Ability of Multimodal Large Language Models [75.45940282834327]
  We introduce Viewpoint Learning, a task designed to evaluate and improve the spatial reasoning capabilities of MLLMs.<n>We present the Viewpoint-100K dataset, consisting of 100K object-centric image pairs with diverse viewpoints and corresponding question-answer pairs.<n>Our approach employs a two-stage fine-tuning strategy, resulting in significant improvements across multiple tasks.
  arXiv  Detail & Related papers  (2025-11-03T14:27:00Z)
• Detect Anything via Next Point Prediction [51.55967987350882]
  Rex- Omni is a 3B-scale MLLM that achieves state-of-the-art object perception performance.<n>On benchmarks like COCO and LVIS, Rex- Omni attains performance comparable to or exceeding regression-based models.
  arXiv  Detail & Related papers  (2025-10-14T17:59:54Z)
• ExACT: Teaching AI Agents to Explore with Reflective-MCTS and Exploratory Learning [78.42927884000673]
  ExACT is an approach to combine test-time search and self-learning to build o1-like models for agentic applications.<n>We first introduce Reflective Monte Carlo Tree Search (R-MCTS), a novel test time algorithm designed to enhance AI agents' ability to explore decision space on the fly.<n>Next, we introduce Exploratory Learning, a novel learning strategy to teach agents to search at inference time without relying on any external search algorithms.
  arXiv  Detail & Related papers  (2024-10-02T21:42:35Z)
• Bidirectional Awareness Induction in Autoregressive Seq2Seq Models [47.82947878753809]
  Bidirectional Awareness Induction (BAI) is a training method that leverages a subset of elements in the network, the Pivots, to perform bidirectional learning without breaking the autoregressive constraints. In particular, we observed an increase of up to 2.4 CIDEr in Image-Captioning, 4.96 BLEU in Neural Machine Translation, and 1.16 ROUGE in Text Summarization compared to the respective baselines.
  arXiv  Detail & Related papers  (2024-08-25T23:46:35Z)
• Towards Automated Identification of Violation Symptoms of Architecture Erosion [2.915855887948474]
  This paper explores the automated identification of violation symptoms from developer discussions in code reviews.<n>We developed 15 machine learning-based classifiers using pre-trained word embeddings and evaluated them on code review comments.<n>Results show that SVM with word2vec achieved the best ML/DL performance with an F1-score of 0.779, while fastText embeddings also yielded strong results.
  arXiv  Detail & Related papers  (2023-06-14T16:20:59Z)
• SSMTL++: Revisiting Self-Supervised Multi-Task Learning for Video Anomaly Detection [108.57862846523858]
  We revisit the self-supervised multi-task learning framework, proposing several updates to the original method. We modernize the 3D convolutional backbone by introducing multi-head self-attention modules. In our attempt to further improve the model, we study additional self-supervised learning tasks, such as predicting segmentation maps.
  arXiv  Detail & Related papers  (2022-07-16T19:25:41Z)
• Anomaly Detection in Video via Self-Supervised and Multi-Task Learning [113.81927544121625]
  Anomaly detection in video is a challenging computer vision problem. In this paper, we approach anomalous event detection in video through self-supervised and multi-task learning at the object level.
  arXiv  Detail & Related papers  (2020-11-15T10:21:28Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.