Research map / Large language models / Retrieval and knowledge
AgentPEN: A Prediction-Explanation Network for Sequential Stock Movement via LLMs and Recurrent Generation
Shuqi Li, Mengyao Guo, Yunzhong Zheng, Siqi Li, Xin Gao, Rui Yan
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add the LLM Text Selection Module (ablation) | AgentPEN · CMIN-US | ACC (%): 53.71 -> 58.09 | Improved |
| use LLM correlation scores instead of random scores for news selection (ablation) | AgentPEN · CMIN-US | ACC (%): 53.61 -> 58.09 | Improved |
| add the Text Memory Module (temporal-spatial memory) (ablation) | AgentPEN · CMIN-US | ACC (%): 52.54 -> 58.09 | Improved |
| add cross-attention Information Fusion Module instead of concatenation (ablation) | AgentPEN · CMIN-US | MCC: 0.030 -> 0.130 | Improved |
| add news text to price-only input (ablation) | AgentPEN · CMIN-US | ACC (%): 53.12 -> 58.09 | Improved |
| add price data to news-only input (ablation) | AgentPEN · CMIN-US | ACC (%): 52.73 -> 58.09 | Improved |
| use stochastic latent variable instead of deterministic mean (ablation) | AgentPEN · AgentPEN-US, CMIN-US, CMIN-CN | MCC: 0.070, 0.076, 0.048 -> 0.118, 0.130, 0.128 | Improved |
| use VAE-based latent generation instead of GRU plus MLP head (ablation) | AgentPEN · AgentPEN-US, CMIN-US, CMIN-CN | MCC: 0.009, 0.010, 0.051 -> 0.118, 0.130, 0.128 | Improved |
| use VAE-based latent generation instead of GRU plus MLP head (ablation) | AgentPEN · CMIN-US | ACC (%): 56.68 -> 58.09 | Improved |
| long-short trading on AgentPEN predictions vs buy-and-hold | AgentPEN · CMIN-CN investment simulation | cumulative return: -5.77% -> +2.13 ± 0.61% | Improved |
| long-only trading on AgentPEN predictions vs buy-and-hold | AgentPEN · CMIN-CN investment simulation | cumulative return: -5.77% -> -3.66 ± 0.99% | Improved |
| AgentPEN vs strongest baseline ExPred | AgentPEN · CMIN-US | ACC (%): 55.85 -> 58.09 | Improved |
Closest papers by meaning
- CausalStock: Deep End-to-end Causal Discovery for News-driven Multi-stock Movement Prediction (NeurIPS 2024 · similarity 0.88)
- MASTER: Market-Guided Stock Transformer for Stock Price Forecasting (AAAI 2024 · similarity 0.86)
- ECHO-GL: Earnings Calls-Driven Heterogeneous Graph Learning for Stock Movement Prediction (AAAI 2024 · similarity 0.86)
- StockMixer: A Simple Yet Strong MLP-Based Architecture for Stock Price Forecasting (AAAI 2024 · similarity 0.85)
- DHMoE: Diffusion Generated Hierarchical Multi-Granular Expertise for Stock Prediction (AAAI 2025 · similarity 0.85)
- FinBen: A Holistic Financial Benchmark for Large Language Models (NeurIPS 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
