aiqbench analysis
CodeTS advances text-to-time-series generation by introducing an executable code intermediate, enabling verifiable synthesis of temporal patterns from natural language descriptions without requiring extensive labeled data. This approach leverages reinforcement learning with verifiable rewards to ensure code executability and output quality, effectively outperforming standard LLM-based baselines on multiple benchmarks. The framework establishes a robust paradigm for generating scenario-specific time series where real-world observations are scarce or expensive to obtain.
CodeTS: Verifiable Text-to-Time Series Generation via Executable Code
arXiv:2609.15393v1 Announce Type: cross Abstract: Text-to-Time Series Generation (Text-to-TS) provides a promising paradigm for synthesizing time series from natural language, enabling scenario-specific generation when real observations are scarce or costly to acquire. However, existing methods typically lack an explicit mechanism for deriving generation logic from textual descriptions to guide time series synthesis. In this paper, we propose CodeTS, a verifiable framework that uses code as an intermediate generation interface, reformulating Text-to-TS generation as a Text-to-Code-to-TS process. CodeTS first maps textual temporal descriptions into an explicit code space, where executable code specifies how textual requirements shape target temporal patterns, and then obtains the time series through code execution. To learn this code generation process reliably without real code annotations, CodeTS constructs aligned Text-Code-TS triplets from structured temporal attributes for supervised initialization. More importantly, we further design multi-stage execution-based rewards that verify format validity, code executability, and time series quality, enabling real Text-TS pairs to provide training signals for Reinforcement Learning with Verifiable Rewards (RLVR). Extensive experiments on eight benchmarks across short, medium, and long generation lengths demonstrate that CodeTS provides a strong zero-shot solution for Text-to-TS generation, outperforming LLM-based baselines and achieving better averaged results than supervised generative baselines trained on the target datasets.
Abstract:Text-to-Time Series Generation (Text-to-TS) provides a promising paradigm for synthesizing time series from natural language, enabling scenario-specific generation when real observations are scarce or costly to acquire. However, existing methods typically lack an explicit mechanism for deriving generation logic from textual descriptions to guide time series synthesis. In this paper, we propose CodeTS, a verifiable framework that uses code as an intermediate generation interface, reformulating Text-to-TS generation as a Text-to-Code-to-TS process. CodeTS first maps textual temporal descriptions into an explicit code space, where executable code specifies how textual requirements shape target temporal patterns, and then obtains the time series through code execution. To learn this code generation process reliably without real code annotations, CodeTS constructs aligned Text-Code-TS triplets from structured temporal attributes for supervised initialization. More importantly, we further design multi-stage execution-based rewards that verify format validity, code executability, and time series quality, enabling real Text-TS pairs to provide training signals for Reinforcement Learning with Verifiable Rewards (RLVR). Extensive experiments on eight benchmarks across short, medium, and long generation lengths demonstrate that CodeTS provides a strong zero-shot solution for Text-to-TS generation, outperforming LLM-based baselines and achieving better averaged results than supervised generative baselines trained on the target datasets.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
arXiv:2609.15393 [cs.LG]
(or arXiv:2609.15393v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2609.15393 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 295 words
Related coverage
Why do models task game?
AI Alignment Forum · Aug 6, 2026
[AINews] Meta Muse Wonder & Wonder Full: MSL’s Claude 5.5-tier Frontier Model
Latent Space · Oct 13, 2026
Performance Profiling on AMD GPUs – Part 5: Profiling-Driven Kernel Optimization with an AI Code-Assist Tool
AMD · Jul 16, 2026
Informe de riesgos de la IA de frontera (febrero–marzo de 2026)
METR · May 19, 2026