T-SMART: Mechanism-Level Attribution for Tool-Augmented Time-Series Question Answering

Ivan Delgado; Himansi Gupta; Bishal Khatri; Niharika Sapre; Lameta Shamoon; Onat Gungor; Tajana Rosing

cs.LGSep 12, 2026

T-SMART: Mechanism-Level Attribution for Tool-Augmented Time-Series Question Answering

Authors: Ivan Delgado, Himansi Gupta, Bishal Khatri, Niharika Sapre, Lameta Shamoon, Onat Gungor, Tajana Rosing

Organizations: University of California, San Diego, CA, USA · West Virginia University, Morgantown, WV, USA

Abstract

Large language models (LLMs) can struggle with time-series question answering (TS-QA), especially when numerical signals are serialized as text and require explicit computation. Tool-augmented approaches improve performance, but existing systems often intertwine language reasoning, computation, and perception, making it difficult to determine which components drive the gains. We present T-SMART, a neurosymbolic framework that separates these roles: a frozen LLM interprets questions and selects operations, deterministic tools perform numerical computation, and structured perception is invoked only when needed. Controlled paired ablations show that deterministic computation provides the dominant benefit, improving accuracy by 31.7 percentage points over direct LLM reasoning on serialized time series, while language understanding and perception offer smaller complementary gains. These results indicate that tool-augmented TS-QA benefits primarily from reliable numerical execution rather than additional language-model reasoning and provide a controlled framework for analyzing component contributions in neurosymbolic time-series systems.

Figures & tables

Fig. 1: The single-pass T-SMART pipeline. One frozen model serves as router, vision sensor, and answer reader; deterministic tools perform every numerical computation, and the visual sensor is invoked only when the trust gate fires.

Branch	Core tools
trend	OLS slope, functional-form fits, ruptures change-points
periodicity	FFT and spectral features, seasonal decomposition
anomaly	z-score and IQR scoring, change-point detection
noise	Engle ARCH-LM, Ljung-Box, ADF, KPSS
similarity	dynamic time warping, statistical distances
causality	Granger F-test, lag correlation

TABLE I: Deterministic branches and the core tools each invokes; the per-branch subtype registry is released with the code.

Name	Setting and role
baseline	trust gate, no loop; A0 numeric head on free-response. Canonical T-SMART; the control condition
vision_override	suppress vision on noise, force anomaly line plots. TimeSeriesExam-selected routing intervention, frozen before OOD testing; a mechanism-transfer ablation
learned_gate	per-branch suppression filter over the trust gate; a lower-look-rate policy fitted on MMTS discordance, evaluated on held-out TimeSeriesExam
vision_off	suppress vision on all branches; the math-only counterfactual
head_grammar , head_planner	A1 / A2 parser over the A0 default; free-response numeric-head variants
llm_only	no router, branch, or tools; one direct answer call. Raw-model control for the fixed-backbone analysis

TABLE II: Configurations used as experimental conditions. A configuration is a named set of pipeline arguments; prompts are identical across conditions.

Model	Setting	OA	PR	NU	AD	SA	CA
Phi-3.5	text	$31.8$	$44$	$24$	$25$	$41$	$25$
Gemini-2.5	text	$30.4$	$29$	$32$	$26$	$35$	$30$
Mistral-7B	text	$34.6$	$29$	$33$	$38$	$45$	$28$
GPT-4o	text	$38.8$	$35$	$32$	$40$	$51$	$36$
ChatTS	TS model	$39.0$	$37$	$39$	$36$	$53$	$30$
Gemini-2.5	vision	$48.8$	$51$	$54$	$46$	$69$	$24$

TABLE III: TimeSeriesExam, per-category accuracy (%). OA is the category-mean overall accuracy.

Model	Micro OA	Base	InWild	Match	Align
ChatTS	$49$	$39$	$50$	$37$	$80$
TS-Agent	$60$	$21$	$71$	$77$	$97$
T-SMART baseline	55.6	$72.7$	$48.3$	$59.0$	$32.9$

TABLE IV: MMTS-Bench, accuracy per subset (%). Micro OA is the row-weighted micro-average. Base uses the numeric head on free-response.

Comparison	$n$	$n_{10}/n_{01}$	$\Delta$	95% CI	$p$
all numerical rows	$393$	128/12	$29.5$	$[+24.4,+34.6]$	$<$ 0.0001
closed-form	$328$	104/0	$31.7$	$[+26.8,+36.9]$	$<$ 0.0001
structured vs. math-only	$568$	113/69	$7.7$	$[+3.2,+12.3]$	0.0014
raw pixels vs. math-only	$568$	119/62	$10.0$	$[+5.5,+14.6]$	$<$ 0.0001
SVI cost (struct. $-$ raw)	$568$		$-2.3$

TABLE V: Paired mechanism effects on identical rows (points). Top: MMTS Base, numeric head vs. llm_compute , scored by Acc@10. Bottom: TimeSeriesExam gated subset, each visual condition vs. math-only.

Layer	A0	A1	A2	A2 fire %
clean	$100$	$100$	$100$	$0$
compositional	$15$	$100$	$100$	$0$
typo	$0$	$83$	$100$	$17$
paraphrase	$0$	$0$	$75$	$100$
overall ( $n{=}240$ )	$29$	$71$	$94$	$29$

TABLE VI: Three-parser study on the synthetic numeric stress set (Acc@10, %; gold = the deterministic tool’s own answer;

n=60

per layer).

Fig. 2: Gated vision helps only where shape and periodicity are decisive: on gate-fired rows (gemini-3.1-flash-lite), structured vision improves periodicity (

+24.8

) and trend (

+14.3

), is near zero on anomaly, and negative on noise (

-4.2

Time Series QA

Time-series question answering connects natural-language questions to evidence and reasoning over sequential data.

See all 32 papers

TimeART: Towards Agentic Time Series Reasoning via Tool-Augmentation
Jan 20, 2026Xingjian Wu, Junkai Lu, Zhengyu Li +5Time Series Reasoning Time Series QA
East China Normal University, Shanghai, China · Aalborg University, Aalborg, Denmark
TS-Skill: A Benchmark for Evaluating Analytical Skills in Time-Series Question Answering
May 23, 2026Liying Han, Kang Yang, Oliver Wang +9LLM Evaluation Time Series Reasoning
University of California, Los Angeles · Samsung Research America · Carnegie Mellon University +2
Beyond Tokenization: Direct Timestep Embedding and Contrastive Alignment for Time-Series Question Answering
Jun 17, 2026Yafeng Wu, Huu Hiep Nguyen, Thin Nguyen +1Time Series Reasoning Time Series QA
Deakin University

Jan 20, 2026J/K move · Enter open · S save

TimeART: Towards Agentic Time Series Reasoning via Tool-Augmentation

Xingjian Wu, Junkai Lu, Zhengyu Li, Xiangfei Qiu+4

East China Normal University, Shanghai, China · Aalborg University, Aalborg, Denmark

Time Series Reasoning Time Series QA Agentic Reasoning LLM Tool Use

Open paper AskPDF arXiv

Time series data widely exist in real-world cyber-physical systems. Though analyzing and interpreting them contributes to significant values, e.g, disaster prediction and financial risk control, current workflows mainly rely on human data scientists, which requires significant labor costs and lacks automation. To tackle this, we introduce TimeART, a framework fusing the analytical capability of strong out-of-the-box tools and the reasoning capability of Large Language Models (LLMs), which serves as a fully agentic data scientist for Time Series Question Answering (TSQA). To teach the LLM-based Time Series Reasoning Models (TSRMs) strategic tool-use, we also collect a 100k expert trajectory corpus called TimeToolBench. To enhance TSRMs' generalization capability, we then devise a four-stage training strategy, which boosts TSRMs through learning from their own early experiences and self-reflections. Experimentally, we train an 8B TSRM on TimeToolBench and equip it with the TimeART framework, and it achieves consistent state-of-the-art performance on multiple TSQA tasks, which pioneers a novel approach towards agentic time series reasoning.

Abstract

Time Series QA

Explore similar work

TimeART: Towards Agentic Time Series Reasoning via Tool-Augmentation

TS-Skill: A Benchmark for Evaluating Analytical Skills in Time-Series Question Answering

Beyond Tokenization: Direct Timestep Embedding and Contrastive Alignment for Time-Series Question Answering