搜索结果：gecko

The ability to use tools is fundamental for large language model (LLM) agents. Given a task, existing systems use LLMs to plan and generate tool calls, which are executed by real-world tools to complete the task. However, tool calls are prone to errors because they are generated primarily from the intrinsic capabilities of LLMs. Moreover, while it is useful to let LLMs iteratively refine the tool-call sequence using execution results from real tools, this process can be expensive and may cause unsafe side effects. To improve LLM tool calls and address issues caused by using real tools for refinement, we introduce Gecko, a stateful simulation environment that provides informative feedback for refining LLM tool calls before real execution. Specifically, Gecko combines rules and LLMs to check the validity of tool names and arguments, synthesize schema-conforming and state-consistent responses, and judge task completion against the user objective. These three types of feedback allow LLMs to refine their tool calls in simulation, forming a simple yet effective test-time scaling method named GATS. On BFCLv3 and $τ^2$-bench, GATS consistently improves the performance of various LLMs.

We map the extraplanar gas, with $\sim$50-200 pc resolution, in nine star-forming galaxies using Multi-Unit Spectroscopic Explorer (MUSE) observations from the GECKOS VLT Large Program targeting edge-on galaxies with similar stellar mass as the Milky Way. The narrow range in stellar mass ($\pm0.35$ dex) of the GECKOS sample makes it ideal for studying trends with star formation rate (SFR). We find strong extraplanar emission reaching $\sim$2-8 kpc from the disk midplane in all targets with $\rm{SFR}\geq$1 M$_{\odot}$ yr$^{-1}$. Targets with SFR$\,\geq\,$5 M$_{\odot}$ yr$^{-1}$ have brighter, more extended H$α$ emission compared to lower SFR targets. In high-SFR systems, the gas velocity dispersion ($σ_{\rm Hα}$) shows a biconical morphology, consistent with the expectation of outflows. This agrees with previous works suggesting high velocity dispersion in a biconical shape is a good means to identify outflows. We find mixed results using line diagnostics ([OIII]$_{5007}$/H$β$ - [NII]/H$α$ and $σ_{\rm Hα}$ - [SII]/H$α$) to spatially resolve ionisation mechanisms across the extraplanar gas. The highest [NII]/H$α$ are the extraplanar gas of the highest SFR systems, yet main-sequence g

Although our lives are increasingly transitioning into the digital world, many digital assets still relate to objects or places in the physical world, e.g., websites of stores or restaurants, digital documents claiming property ownership, or digital identifiers encoded in QR codes for mobile payments in shops. Currently, users cannot securely associate digital assets with their related physical space, leading to problems such as fake brand stores, property fraud, and mobile payment scams. In many cases, the necessary information to protect digital assets exists, e.g., via contractual relationships and cadaster entries, but there is currently no uniform way of retrieving and verifying these documents. In this work, we propose the Geo-Enabled Cryptographic Key Oracle (GECKO), a geographical PKI that provides a global view of digital assets based on their geo-location and occupied space. GECKO allows for the bidirectional translation of trust between the physical and digital world. Users can verify which assets are supposed to exist at their location, as well as verify which physical space is claimed by a digital entity. GECKO supplements current PKI systems and can be used in additio

We present a spatially resolved, multiphase study of the outflow in the edge-on starburst galaxy ESO~484-036 from the GECKOS survey, combining VLT/MUSE H$α$ and ALMA CO(1$-$0) observations to analyse the atomic ionised and cold molecular gas. Both show extraplanar emission consistent with a conical outflow. Ionised gas is enclosed by molecular gas, which is detected up to 2.5 kpc from the disc. Molecular gas dominates near the disc, except at the nuclear base, while ionised gas extends beyond 3 kpc. The deprojected outflow velocities are $\lesssim400\ \rm km\ s^{-1}$ in both phases and are consistent with ballistic motion, with some gas possibly falling back onto the disc. We find that the mass outflow rates are in the range of $\dot M_{\rm ion}\sim1-5\ \rm M_\odot\ \rm yr^{-1}$ and $\dot M_{\rm mol}\sim13-54\ \rm M_\odot\ \rm yr^{-1}$, giving mass loading factors of $η_{M\rm, ion}\sim 0.1-0.6$ and $η_{M\rm, mol}\sim 1.5-6.2$. These ranges reflect velocity and geometric uncertainties. Despite the short depletion time ($τ_{\rm dep} = 16-48\rm\ Myr$), the outflow may regulate rather than permanently quench the gas reservoir. Energy loading ($η_E\leq0.16$) and momentum loading ($η_p\l

We present Gecko, a compact and versatile text embedding model. Gecko achieves strong retrieval performance by leveraging a key idea: distilling knowledge from large language models (LLMs) into a retriever. Our two-step distillation process begins with generating diverse, synthetic paired data using an LLM. Next, we further refine the data quality by retrieving a set of candidate passages for each query, and relabeling the positive and hard negative passages using the same LLM. The effectiveness of our approach is demonstrated by the compactness of the Gecko. On the Massive Text Embedding Benchmark (MTEB), Gecko with 256 embedding dimensions outperforms all existing entries with 768 embedding size. Gecko with 768 embedding dimensions achieves an average score of 66.31, competing with 7x larger models and 5x higher dimensional embeddings.

We present a multiphase, resolved study of the galactic wind extending from the nearby starburst galaxy NGC 4666. For this we use VLT/MUSE observations from the GECKOS program and HI data from the WALLABY survey. We identify both ionised and HI gas in a biconical structure extending to at least $z\sim$8 kpc from the galaxy disk, with increasing velocity offsets above the midplane in both phases, consistent with a multiphase wind. The measured electron density, using [SII], differs significantly from standard expectations of galactic winds. We find electron density declines from the galaxy centre to $\sim2$ kpc, then rises again, remaining high ($\sim100-300$ cm$^{-3}$) out to $\sim$5 kpc. We find that HI dominates the mass loading. The total HI mass outflow rate (above $z~>2$ kpc) is between $5-13~M_{\odot}~\rm yr^{-1}$, accounting for uncertainties from disk-blurring and group interactions. The total ionised mass outflow rate (traced by H$α$) is between $0.5~M_{\odot}~\rm yr^{-1}$ and $5~M_{\odot}~\rm yr^{-1}$, depending on $n_e(z)$ assumptions. From ALMA/ACA observations, we place an upper-limit on CO flux in the outflow which correlates to $\lesssim2.9~M_{\odot}~\rm yr^{-1}$.

We introduce GECKO, a bilingual large language model (LLM) optimized for Korean and English, along with programming languages. GECKO is pretrained on the balanced, high-quality corpus of Korean and English employing LLaMA architecture. In this report, we share the experiences of several efforts to build a better data pipeline for the corpus and to train our model. GECKO shows great efficiency in token generations for both Korean and English, despite its small size of vocabulary. We measure the performance on the representative benchmarks in terms of Korean, English and Code, and it exhibits great performance on KMMLU (Korean MMLU) and modest performance in English and Code, even with its smaller number of trained tokens compared to English-focused LLMs. GECKO is available to the open-source community under a permissive license. We hope our work offers a research baseline and practical insights for Korean LLM research. The model can be found at: https://huggingface.co/kifai/GECKO-7B

Disentangling the (co-)evolution of individual galaxy structural components remains a difficult task, owing to the inability to cleanly isolate light from spatially overlapping components. In this pilot study of PGC\,044931, observed as part of the GECKOS survey, we utilise a VIRCAM $H$-band image to decompose the galaxy into five photometric components, three of which dominate by contributing $>50\%$ of light in specific regions: a main disc, a boxy/peanut bulge, and a nuclear disc. When the photometric decompositions are mapped onto MUSE observations, we find remarkably good separation in stellar kinematic space. All three structures occupy unique locations in the parameter space of the ratio of the light-weighted stellar line-of-sight mean velocity and velocity dispersion ($\rm{V}_{\star}/σ_{\star}$), and the high-order stellar skew ($h_{3}$). These clear and distinct kinematic behaviours allow us to make inferences about the formation histories of the individual components from observations of the mean stellar ages and metallicities of the three components. A clear story emerges: the main disc built over a sustained and extended star formation phase, possibly partly fuelled

The gravitational wave (GW) event S230518h is a potential binary neutron star-black hole merger (NSBH) event that was detected during engineering run 15 (ER15), which served as the commissioning period before the LIGO-Virgo-KAGRA (LVK) O4a observing run. Despite its low probability of producing detectable electromagnetic emissions, we performed extensive follow-up observations of this event using the GECKO telescopes in the southern hemisphere. Our observation covered 61.7\% of the 90\% credible region, a $\rm 284\:deg^2$ area accessible from the southern hemisphere, reaching a median limiting magnitude of $R=21.6$ mag. In these images, we conducted a systematic search for an optical counterpart of this event by combining a CNN-based classifier and human verification. We identified 128 transient candidates, but no significant optical counterpart was found that could have caused the GW signal. Furthermore, we provide feasible KN properties that are consistent with the upper limits of observation. Although no optical counterpart was found, our result demonstrates both GECKO's efficient wide-field follow-up capabilities and usefulness for constraining properties of kilonovae from NSBH

The central regions of disc galaxies host a rich variety of stellar structures: nuclear discs, bars, bulges, and boxy-peanut (BP) bulges. These components are often difficult to disentangle, both photometrically and kinematically, particularly in star-forming galaxies where dust obscuration and complex stellar motions complicate interpretation. In this work, we use data from the GECKOS-MUSE survey to investigate the impact of dust on axisymmetric Jeans Anisotropic Multi-Gaussian Expansion (JAM) models, and assess their ability to recover kinematic structure in edge-on disc galaxies. We construct JAM models for a sample of seven edge-on ($i \gtrapprox 85^\circ$) galaxies that span a range of star formation rates, dust content, and kinematic complexity. We find that when dust is appropriately masked, the disc regions of each galaxy are fit to $χ^2_{\text{reduced}}\leq 5$. We analyse two-dimensional residual velocity fields to identify signatures of non-axisymmetric structure. We find that derived dynamical masses are constant within 10% for each galaxy across all dust masking levels. In NGC 3957, a barred boxy galaxy in our sample, we identify velocity residuals that persist even und

Pretraining a Multiple Instance Learning (MIL) aggregator enables the derivation of Whole Slide Image (WSI)-level embeddings from patch-level representations without supervision. While recent multimodal MIL pretraining approaches leveraging auxiliary modalities have demonstrated performance gains over unimodal WSI pretraining, the acquisition of these additional modalities necessitates extensive clinical profiling. This requirement increases costs and limits scalability in existing WSI datasets lacking such paired modalities. To address this, we propose Gigapixel Vision-Concept Knowledge Contrastive pretraining (GECKO), which aligns WSIs with a Concept Prior derived from the available WSIs. First, we derive an inherently interpretable concept prior by computing the similarity between each WSI patch and textual descriptions of predefined pathology concepts. GECKO then employs a dual-branch MIL network: one branch aggregates patch embeddings into a WSI-level deep embedding, while the other aggregates the concept prior into a corresponding WSI-level concept embedding. Both aggregated embeddings are aligned using a contrastive objective, thereby pretraining the entire dual-branch MIL m

We present GECKOS (Generalising Edge-on galaxies and their Chemical bimodalities, Kinematics, and Outflows out to Solar environments), a new ESO VLT/MUSE large program. The main aim of GECKOS is to reveal the variation in key physical processes of disk formation by connecting Galactic Archaeology with integral field spectroscopic observations of nearby galaxies. Edge-on galaxies are ideal for this task: they allow us to disentangle the assembly history imprinted in thick disks and provide the greatest insights into outflows. The GECKOS sample of 35 nearby edge-on disk galaxies is designed to trace the assembly histories and properties of galaxies across a large range of star formation rates, bulge-to-total ratios, and boxy and non-boxy bulges. GECKOS will deliver spatially resolved measurements of stellar abundances, ages, and kinematics, as well as ionised gas metallicities, ionisation parameters, pressure, and inflow and outflow kinematics; all key parameters for building a complete chemodynamical picture of disk galaxies. With these data, we aim to extend Galactic analysis methods to the wider galaxy population, reaping the benefits of detailed Milky Way studies, while probing t

The vertical evolution of galactic discs is governed by the sub-structures within them. We examine the diversity of kinematic sub-structure present in the first 12 galaxies observed from the GECKOS survey, a VLT/MUSE large programme providing a systematic study of 36 edge-on, Milky Way-mass disc galaxies. Employing the nGIST analysis pipeline, we derive the mean line-of-sight stellar velocity ($V_{\star}$), velocity dispersion ($σ_{\star}$), skew ($h_{3}$), and kurtosis ($h_{4}$) for the sample, and examine 2D maps and 1D line profiles. Visually, the majority of this sample (8/12) are found to possess boxy-peanut bulges and host the corresponding kinematic structure predicted for stellar bars viewed in projection. Four galaxies exhibit strong evidence for the presence of nuclear discs, including central $h_{3}$-$V_{\star}$ sign mismatch, `croissant'-shaped central depressions in $σ_{\star}$ maps, strong gradients in $h_{3}$, and positive $h_{4}$ plateaus over the expected nuclear disc extent. The strength of the $h_{3}$ feature corresponds to the size of the nuclear disc, measured from the $h_{3}$ turnover radius. We can explain the features within the kinematic maps of all sample

While text-to-image (T2I) generative models have become ubiquitous, they do not necessarily generate images that align with a given prompt. While previous work has evaluated T2I alignment by proposing metrics, benchmarks, and templates for collecting human judgements, the quality of these components is not systematically measured. Human-rated prompt sets are generally small and the reliability of the ratings -- and thereby the prompt set used to compare models -- is not evaluated. We address this gap by performing an extensive study evaluating auto-eval metrics and human templates. We provide three main contributions: (1) We introduce a comprehensive skills-based benchmark that can discriminate models across different human templates. This skills-based benchmark categorises prompts into sub-skills, allowing a practitioner to pinpoint not only which skills are challenging, but at what level of complexity a skill becomes challenging. (2) We gather human ratings across four templates and four T2I models for a total of >100K annotations. This allows us to understand where differences arise due to inherent ambiguity in the prompt and where they arise due to differences in metric and

One of the keys to the success of multimessenger astronomy is the rapid identification of the electromagnetic wave counterpart, kilonova (KN), of the gravitational-wave (GW) event. Despite its importance, it is hard to find a KN associated with a GW event, due to a poorly constrained GW localization map and numerous signals that could be confused as a KN. Here, we present the Gravitational-wave Electromagnetic wave Counterpart Korean Observatory (GECKO) project, the GECKO observation of GW190425, and prospects of GECKO in the fourth observing run (O4) of the GW detectors. We outline our follow-up observation strategies during O3. In particular, we describe our galaxy-targeted observation criteria that prioritize based on galaxy properties. Armed with this strategy, we performed an optical and/or near-infrared follow-up observation of GW190425, the first binary neutron star merger event during the O3 run. Despite a vast localization area of 7460 deg^2, we observed 621 host galaxy candidates, corresponding to 29.5% of the scores we assigned, with most of them observed within the first 3 days of the GW event. Ten transients were discovered during this search, including a new transient

Gravitational forces can induce deviations in body posture from desired configurations in multi-legged arboreal robot locomotion with low leg stiffness, affecting the contact angle between the swing leg's end-effector and the climbing surface during the gait cycle. The relationship between desired and actual foot positions is investigated here in a leg-stiffness-enhanced model under external forces, focusing on the challenge of unreliable end-effector attachment on climbing surfaces in such robots. Inspired by the difference in ceiling attachment postures of dead and living geckos, feedforward compensation of the stance phase legs is the key to solving this problem. A feedforward gravity compensation (FGC) strategy, complemented by leg coordination, is proposed to correct gravity-influenced body posture and improve adhesion stability by reducing body inclination. The efficacy of this strategy is validated using a quadrupedal climbing robot, EF-I, as the experimental platform. Experimental validation on an inverted surface (ceiling walking) highlight the benefits of the FGC strategy, demonstrating its role in enhancing stability and ensuring reliable end-effector attachment without

Adaptive structures are of interest for their ability to dynamically modify mechanical properties post fabrication, enabling structural performance that is responsive to environmental uncertainty and changing loading conditions. Dynamic control of stiffness is of particular importance as a fundamental structural property, impacting both static and dynamic structural performance. However, existing technologies necessitate continuous power to maintain multiple stiffness states or couple stiffness modulation to a large geometric reconfiguration. In this work, reversible lamination of stiff materials using Gecko-inspired dry adhesives is leveraged for bending stiffness control. All stiffness states are passively maintained, with electrostatic or magnetic actuation applied for ~1s to reprogram stiffness. We demonstrate hinges with up to four passively maintained reprogrammable states decoupled from any shape reconfiguration. Design guidelines are developed for maximizing stiffness modulation. Experimentally, the proposed method achieved a stiffness modulation ratio of up to 14.4, with simulations showing stiffness modulation ratios of at least 73.0. It is anticipated that the stiffness