# TIME: Temporally Intelligent Meta-reasoning Engine for Context-Triggered Explicit Reasoning

arXiv: https://arxiv.org/abs/2601.05300 · HTML: https://mlsanity.com/paper/2601.05300/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Knowledge- and tool-augmented reasoning, one of 17 directions in LLM reasoning.
- Builds on 7 of its 20 references that are in the map.
- First posted on arXiv 8 Jan 2026.

## Placed in

- LLM reasoning / [Knowledge- and tool-augmented reasoning](https://mlsanity.com/lr/knowledge-and-tool-augmented-reasoning/)

## Builds on (most cited first, 7 of 7)

- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Reasoning Models Don't Always Say What They Think](https://mlsanity.com/paper/2505.05410/) (2025)
- [L1: Controlling How Long A Reasoning Model Thinks With Reinforcement Learning](https://mlsanity.com/paper/2503.04697/) (2025)
- [Interleaved Reasoning for Large Language Models via Reinforcement Learning](https://mlsanity.com/paper/2505.19640/) (2025)
- [DOTS: Learning to Reason Dynamically in LLMs via Optimal Reasoning Trajectories Search](https://mlsanity.com/paper/2410.03864/) (2024)
- [Time Awareness in Large Language Models: Benchmarking Fact Recall Across Time](https://mlsanity.com/paper/2409.13338/) (2024)
- [RAG Meets Temporal Graphs: Time-Sensitive Modeling and Retrieval for Evolving Knowledge](https://mlsanity.com/paper/2510.13590/) (2025)

## Built on by (most cited first, 1 of 1)

- [CoT2-Meta: Budgeted Metacognitive Control for Test-Time Reasoning](https://mlsanity.com/paper/2603.28135/) (2026)

## Abstract (arXiv)

Reasoning-oriented language models typically expose explicit reasoning as a long, front-loaded chain of "thinking" tokens before the main output, either always enabled or externally toggled at inference time. Although this can help on arithmetic, coding, and other multi-step tasks, it is costly, weakens claim-level auditability, and does not allow the model to re-trigger explicit reasoning once presentation has begun. In dialogue, these limitations are compounded by weak sensitivity to temporal structure: unless time is explicitly stated in text, standard models treat replies separated by seconds and replies separated by weeks as equivalent. We introduce TIME (Temporally Intelligent Meta-reasoning Engine), a behavioral alignment framework that learns explicit reasoning as a context-triggered control policy rather than a fixed response mode. TIME augments dialogue with optional ISO 8601 <time> tags, tick events that represent silent time passage, and short <think> blocks that may appear anywhere in a response. Using a four-phase curriculum, including a small maximally diverse full-batch alignment stage, we train Qwen3 dense models to invoke brief, in-place reasoning bursts only when contextual cues warrant them, while keeping user-facing output compact. We also introduce TIMEBench, a diagnostic benchmark for evaluating reasoning from temporal cues in dialogue. Across 4B-32B scales, TIME improves TIMEBench scores over the corresponding base Qwen3 models in both thinking and no-thinking modes while reducing explicit reasoning tokens by roughly an order of magnitude. Beyond score improvements, TIME induces a distinct behavioral shift: explicit reasoning becomes more compact and more responsive to contextual cues. Code, training data, and benchmark artifacts are publicly available.