Large Reasoning Models (LRMs) achieve impressive performance on complex reasoning tasks via Chain-of-Thought (CoT) reasoning, which enables them to generate intermediate thinking tokens before arriving at the final answer. However, LRMs often suffer from significant overthinking, spending excessive compute time even after the answer is generated early on. Prior work has identified the existence of an optimal reasoning length such that truncating reasoning at this point significantly shortens CoT outputs with virtually no change in performance. However, determining optimal CoT lengths for practical datasets is highly non-trivial as they are fully task and model-dependent. In this paper, we precisely address this and design Terminator, an early-exit strategy for LRMs at inference to mitigate overthinking. The central idea underpinning Terminator is that the first arrival of an LRM's final answer is often predictable, and we leverage these first answer positions to create a novel dataset of optimal reasoning lengths to train Terminator. Powered by this approach, Terminator achieves significant reductions in CoT lengths of 14%-55% on average across four challenging practical datasets:
Transmission spectroscopy has been a workhorse technique over the past two decades to constrain the physical and chemical properties of exoplanet atmospheres. One of its classical key assumptions is that the portion of the atmosphere it probes -- the terminator region -- is homogeneous. Several works in the past decade, however, have put this into question for highly irradiated, hot ($T_{eq}\gtrsim 1000$ K) gas giant exoplanets both empirically and via 3-dimensional modelling. While models predict clear differences between the evening (day-to-night) and morning (night-to-day) terminators, direct morning/evening transmission spectra in a wide wavelength range has not been reported for an exoplanet to date. Under the assumption of precise and accurate orbital parameters on WASP-39 b, here we report the detection of inhomogeneous terminators on the exoplanet WASP-39 b, which allows us to retrieve its morning and evening transmission spectra in the near-infrared ($2-5\ μ$m) using JWST. We observe larger transit depths in the evening which are, on average, $405 \pm 88$ ppm larger than the morning ones, also having qualitatively larger features than the morning spectrum. The spectra are
The perturbations from the solar terminator in the range of acoustic-gravity waves (AGWs) periods from 5 minutes to 1 hour were analysed with the use of measurements of VLF radio signals amplitudes on the European radio path GQD--A118 (Great Britain--France). These observations provide information on the propagation of waves at altitudes near the mesopause ($\sim$ 90 km), where VLF radio waves are reflected. On the considered radio path a systematic increase in fluctuations in the amplitudes of radio waves was observed within a few hours after the passage of the evening terminator. For April, June, October 2020 and February 2021 events, the distribution of the number of wave perturbations with large amplitudes over AGWs time periods has been studied. Our results show that the evening terminator for different seasons is dominated by waves in the range of periods of 15--20 minutes. The amplitudes of the AGWs from the terminator at the heights of the mesosphere (fluctuations in the concentration of neutral particles, velocity components and vertical displacement of the volume element) are approximately determined by the fluctuations of the amplitudes of the VLF radio signals. The ampl
Consider a setting where a pre-trained agent is operating in an environment and a human operator can decide to temporarily terminate its operation and take-over for some duration of time. These kind of scenarios are common in human-machine interactions, for example in autonomous driving, factory automation and healthcare. In these settings, we typically observe a trade-off between two extreme cases -- if no take-overs are allowed, then the agent might employ a sub-optimal, possibly dangerous policy. Alternatively, if there are too many take-overs, then the human has no confidence in the agent, greatly limiting its usefulness. In this paper, we formalize this setup and propose an explainability scheme to help optimize the number of human interventions.
We previously identified an event in the solar timeline that appeared to play a role in how Sunspot Cycle 23 (SC23) transitioned into Sunspot Cycle 24 (SC24). The timeframe for this transition was rapid, taking place over a very short time and perhaps in a time as short as a single solar rotation. Further, we inferred that the transition observed was a critical moment for the Sun's global-scale magnetic field as it was being manifest in the spatially and temporally overlapping magnetic systems belonging to the Sun's 22-year (Hale) magnetic cycle. These events have been dubbed as Hale Cycle terminations, or `terminators' for short. Subsequent exploration of the sunspot record revealed a relationship between terminator separation (as a measure of overlap in the Hale Cycles) and the upcoming sunspot cycle amplitude using a Hilbert transform. Finally, we extrapolated the contemporary sunspots data's Hilbert phase function to project the occurrence of the SC24 terminator in Mid-2020 and inferred that this would result in a large Sunspot Cycle 25 (SC25) amplitude. This paper presents observational evidence that the end of SC24 and the initial growth of SC25 followed a terminator that occ
Recent research has demonstrated the existence of a new type of solar event, the "terminator." Unlike the Sun's signature events, flares and Coronal Mass Ejections, the terminator most likely originates in the solar interior, at or near the tachocline. The terminator signals the end of a magnetic activity cycle at the Sun's equator and the start of a sunspot cycle at mid latitudes. Observations indicate that the time difference between these events is very short, less than a solar rotation, in the context of the sunspot cycle. As the (definitive) start and end point of solar activity cycles the precise timing of terminators should permit new investigations into the meteorology of our star's atmosphere. In this letter we use a standard method in signal processing, the Hilbert transform, to identify a mathematically robust signature of terminators in sunspot records and in radiative proxies. Using a linear extrapolation of the Hilbert phase of the sunspot number and F10.7 solar radio flux time series we can achieve higher fidelity historical terminator timing than previous estimates have permitted. Further, this method presents a unique opportunity to project, from analysis of sunspo
Rocky planets orbiting M-dwarf stars are among the most promising and abundant astronomical targets for detecting habitable climates. Planets in the M-dwarf habitable zone are likely synchronously rotating, such that we expect significant day-night temperature differences, and potentially limited fractional habitability. Previous studies have focused on scenarios where fractional habitability is confined to the substellar or "eye" region, but in this paper we explore the possibility of planets with terminator habitability, defined by the existence of a habitable band at the transition between a scorching dayside and a glacial nightside. Using a global climate model, we show that for water-limited planets it is possible to have scorching temperatures in the "eye" and freezing temperatures on the nightside, while maintaining a temperate climate in the terminator region, due to a reduced atmospheric energy transport. Whereas on water-rich planets, increasing stellar flux leads to increased atmospheric energy transport and a reduction in day-night temperature differences, such that the terminator does not remain habitable once the dayside temperatures approach runaway or moist greenhou
This paper is devoted to the extraction of entities and semantic relations between them from scientific texts, where we consider scientific terms as entities. In this paper, we present a dataset that includes annotations for two tasks and develop a system called TERMinator for the study of the influence of language models on term recognition and comparison of different approaches for relation extraction. Experiments show that language models pre-trained on the target language are not always show the best performance. Also adding some heuristic approaches may improve the overall quality of the particular task. The developed tool and the annotated corpus are publicly available at https://github.com/iis-research-team/terminator and may be useful for other researchers.
Computational protein design has the potential to deliver novel molecular structures, binders, and catalysts for myriad applications. Recent neural graph-based models that use backbone coordinate-derived features show exceptional performance on native sequence recovery tasks and are promising frameworks for design. A statistical framework for modeling protein sequence landscapes using Tertiary Motifs (TERMs), compact units of recurring structure in proteins, has also demonstrated good performance on protein design tasks. In this work, we investigate the use of TERM-derived data as features in neural protein design frameworks. Our graph-based architecture, TERMinator, incorporates TERM-based and coordinate-based information and outputs a Potts model over sequence space. TERMinator outperforms state-of-the-art models on native sequence recovery tasks, suggesting that utilizing TERM-based and coordinate-based features together is beneficial for protein design.
In this work we study the dynamics in an apocalypse where the individuals can temporarily become zombies, returning to the living state again or die. We describe this dynamics using a modified version of the epidemic SIS model. The zombies can die when in contact with a terminator, which is part of the living people. To define the possible interactions we use an Erdös-Rènyi network and we calculate how the absorbing and active phase of the original SIS model is influenced by the average network degree. We also use the component and domain distribution of the network to understand how this influence occurs.
The complexity of atmospheric retrieval models is largely data-driven and one-dimensional models have generally been considered adequate with current data quality. However, recent studies have suggested that using 1D models in retrievals can result in anomalously cool terminator temperatures and biased abundance estimates even with existing transmission spectra of hot Jupiters. Motivated by these claims and upcoming high-quality transmission spectra we systematically explore the limitations of 1D models using synthetic and current observations. We use 1D models of varying complexity, both analytic and numerical, to revisit claims of biases when interpreting transmission spectra of hot Jupiters with inhomogeneous terminator compositions. Overall, we find the reported biases to be resulting from specific model assumptions rather than intrinsic limitations of 1D atmospheric models in retrieving current observations of asymmetric terminators. Additionally, we revise atmospheric retrievals of the hot Jupiter WASP-43b ($T_{\rm eq}=1440$ K) and the ultra-hot Jupiter WASP-103b ($T_{\rm eq}=2484$ K ) for which previous studies inferred abnormally cool atmospheric temperatures. We retrieve t
We present the problem of reinforcement learning with exogenous termination. We define the Termination Markov Decision Process (TerMDP), an extension of the MDP framework, in which episodes may be interrupted by an external non-Markovian observer. This formulation accounts for numerous real-world situations, such as a human interrupting an autonomous driving agent for reasons of discomfort. We learn the parameters of the TerMDP and leverage the structure of the estimation problem to provide state-wise confidence bounds. We use these to construct a provably-efficient algorithm, which accounts for termination, and bound its regret. Motivated by our theoretical analysis, we design and implement a scalable approach, which combines optimism (w.r.t. termination) and a dynamic discount factor, incorporating the termination probability. We deploy our method on high-dimensional driving and MinAtar benchmarks. Additionally, we test our approach on human data in a driving setting. Our results demonstrate fast convergence and significant improvement over various baseline approaches.
Proving termination or non-termination of lasso programs is a challenging problem in program verification. To unify state-of-the-art approaches under a common execution framework, we present PaSTTeL, a modular and generic parallel portfolio framework for termination and non-termination analysis of lasso programs. PaSTTeL is designed to: (1) facilitate the integration of new analysis algorithms into the portfolio, (2) execute registered strategies concurrently, and (3) act as a self-contained library component that can be seamlessly embedded into any external project requiring (non-)termination analysis. Initial experiments demonstrate that an instantiation of PaSTTeL performs competitively against state-of-the-art tools.
It has been shown that, regarding a terminating right-linear overlay term rewrite system (TRS), any rewrite sequence terminating in a normal form can be simulated by an innermost reduction. In this paper, using this simulation property, we show that for a right-linear overlay TRS, there is no infinite minimal dependency-pair chain if and only if there is no infinite innermost minimal dependency-pair chain. As a consequence, termination and innermost termination coincide for the class of right-linear overlay TRSs.
Mastering terminal environments requires language agents capable of multi-step planning, feedback-grounded execution, and dynamic state adaptation. However, training such agents is currently bottlenecked by a reliance on scraped external repositories, which limits domain diversity, environment controllability, and the targeting of specific capability deficits. We introduce LiteCoder-Terminal-Gen, a zero-dependency synthesis pipeline that autonomously generates executable and verifiable terminal training environments directly from domain specifications. Using this framework, we construct two large-scale resources: LiteCoder-Terminal-SFT, comprising 11,255 expert trajectories across 10 domains, and LiteCoder-Terminal-RL, featuring 602 verifiable environments for trajectory-level preference optimization. Supervised fine-tuning of Qwen-family models on our SFT dataset yields agents that significantly outperform their base counterparts. Notably, our 32B variant achieves 29.06%, 18.54%, and 34.00% pass@1 on Terminal Bench 1.0, 2.0, and Pro, respectively. Furthermore, applying Direct Multi-turn Preference Optimization (DMPO) on our RL environments yields additional performance gains. Thes
Environments are the bottleneck for self-improving agents. Current terminal benchmarks were built for evaluation, not training; reinforcement learning requires a scalable pipeline, not just a dataset. We introduce Endless Terminals, a fully autonomous pipeline that procedurally generates terminal-use tasks without human annotation. The pipeline has four stages: generating diverse task descriptions, building and validating containerized environments, producing completion tests, and filtering for solvability. From this pipeline we obtain 3255 tasks spanning file operations, log management, data processing, scripting, and database operations. We train agents using vanilla PPO with binary episode level rewards and a minimal interaction loop: no retrieval, multi-agent coordination, or specialized tools. Despite this simplicity, models trained on Endless Terminals show substantial gains: on our held-out dev set, Llama-3.2-3B improves from 4.0% to 18.2%, Qwen2.5-7B from 10.7% to 53.3%, and Qwen3-8B-openthinker-sft from 42.6% to 59.0%. These improvements transfer to human-curated benchmarks: models trained on Endless Terminals show substantial gains on held out human curated benchmarks: on
Terminal agents extend Large Language Models with the ability to execute tasks directly in command-line environments, but their progress is bottlenecked by the scarcity of high-quality training data. Existing approaches bootstrap from partial sources such as human-defined seeds or GitHub repositories to instantiate one component and then complete the rest, producing tasks confined to narrow seed distributions, environments misaligned with task semantics, and inefficient trajectories from unguided exploration. To address these limitations, we introduce Terminal-World, a fully automated pipeline that uses agent skills as the central synthesis primitive, which jointly encode what to accomplish, when to apply (preconditions and environment state), and how to execute, enabling task instructions, environments, and teacher trajectories to be co-derived. To further broaden the synthesis space, Terminal-World composes skills into skill teams and skill graphs for multi-role and cross-domain task synthesis. Using this pipeline, we construct 5,723 training environments and train Terminal-World-8B/14B/32B, evaluated across 6 benchmarks where the Terminal-World series consistently outperforms te
Determining whether a program terminates is a core challenge in program analysis with direct implications for correctness, verification, and security. We investigate whether transformer architectures can recognise termination patterns directly from source code and how their strengths can be amplified through ensembles. To overcome the extreme scarcity of non-terminating examples, we design an ensemble framework of compact transformer encoders, systematically trained with a suite of imbalance-aware loss functions and class-aware sampling techniques. By combining models trained with distinct loss functions, our ensembles achieve substantially stronger performance than any single transformer, outperforming both powerful off-the-shelf LLMs and graph-based methods. Finally, we introduce an attribution pipeline that produces syntax-aware explanations for the termination estimation.
Analysis of data from randomized controlled trials in vulnerable populations requires special attention when assessing treatment effect by a score measuring, e.g., disease stage or activity together with onset of prevalent terminal events. In reality, it is impossible to disentangle a disease score from the terminal event, since the score is not clinically meaningful after this event. In this work, we propose to assess treatment interventions simultaneously on the terminal event and the disease score in the absence of a terminal event. Our proposal is based on a natural data-generating mechanism, respecting that a disease score does not exist beyond the terminal event. We use modern semi-parametric statistical methods to provide robust and efficient estimation of the risk of terminal event and expected disease score conditional on no terminal event at a pre-specified landmark time. We also use the simultaneous asymptotic behaviour of our estimators to develop a powerful closed testing procedure for confirmatory assessment of treatment effect on both onset of terminal event and level of disease score in the absence of a terminal event. A simulation study mimicking a large-scale outc
The Unix terminal, or just simply, the terminal, can be found being applied in almost every facet of computing. It is available across all major platforms and often integrated into other applications. Due to its ubiquity, even marginal improvements to the terminal have the potential to make massive improvements to productivity on a global scale. We believe that evolutionary improvements to the terminal, in its current incarnation as windowed terminal emulator, are possible and that developing a thorough understanding of issues that current terminal users face is fundamental to knowing how the terminal should evolve. In order to develop that understanding we have mined Unix and Linux Stack Exchange using a fully-reproducible method which was able to extract and categorize 91.0% of 1,489 terminal-related questions (from the full set of nearly 240,000 questions) without manual intervention. We present an analysis, to our knowledge the first of its kind, of windowed terminal-related questions posted over a 15-year period and viewed, in aggregate, approximately 40 million times. As expected, given its longevity, we find the terminal's many features being applied across a wide variety of