SPT-CL J0417-4748: A Deep Chandra Study of a Relaxed Galaxy Cluster Without Central Star Formation

Kavli Affiliate: Steven Allen | Summary:We present an in-depth Chandra X-ray analysis of the galaxy cluster SPT-CL J0417$-$4748 (hereafter SPT J0417), at z = 0.58, with a focus on its thermodynamic properties and the apparent absence of central star formation. Utilizing a total Chandra exposure of 103 ks, we find that the large-scale X-ray morphology […]


Continue.. SPT-CL J0417-4748: A Deep Chandra Study of a Relaxed Galaxy Cluster Without Central Star Formation

Preserve-Then-Quantize: Balancing Rank Budgets for Quantization Error Reconstruction in LLMs

Kavli Affiliate: Hsiao-Mei (Sherry) Cho| First 5 Authors: [#item_custom_name[1, [#item_custom_name[2, [#item_custom_name[3, [#item_custom_name[4, [#item_custom_name[5| Summary:Quantization Error Reconstruction (QER) reduces accuracy loss in Post-Training Quantization (PTQ) by approximating weights as $mathbfW approx mathbfQ + mathbfLmathbfR$, using a rank-$r$ correction to reconstruct quantization error. Prior methods devote the full rank budget to error reconstruction, which is suboptimal when […]


Continue.. Preserve-Then-Quantize: Balancing Rank Budgets for Quantization Error Reconstruction in LLMs

Enhancing Multi-Image Understanding through Delimiter Token Scaling

Kavli Affiliate: Hsiao-Mei (Sherry) Cho| First 5 Authors: [#item_custom_name[1, [#item_custom_name[2, [#item_custom_name[3, [#item_custom_name[4, [#item_custom_name[5| Summary:Large Vision-Language Models (LVLMs) achieve strong performance on single-image tasks, but their performance declines when multiple images are provided as input. One major reason is the cross-image information leakage, where the model struggles to distinguish information across different images. Existing LVLMs already […]


Continue.. Enhancing Multi-Image Understanding through Delimiter Token Scaling

The Effect of Tidal Heating and Volatile Budgets on the Outgassed Atmosphere of 55 Cancri e

Kavli Affiliate: Laura Schaefer | Summary:55 Cancri e is a $sim$8 Gyr rocky world (1.95 $R_oplus$, 8.8 $M_oplus$) orbiting a K-type star. JWST observations suggest a carbon-dominated atmosphere (CO$_2$/CO) over a global magma ocean ($>$3000 K). We suggest that any CO$_2$-dominated atmosphere, with trace H$_2$O/O$_2$, likely arises from outgassing of its initial volatile reservoir. As […]


Continue.. The Effect of Tidal Heating and Volatile Budgets on the Outgassed Atmosphere of 55 Cancri e

Toward Pluralizing Reflection in HCI through Daoism

Kavli Affiliate: Hsiao-Mei (Sherry) Cho| First 5 Authors: [#item_custom_name[1, [#item_custom_name[2, [#item_custom_name[3, [#item_custom_name[4, [#item_custom_name[5| Summary:Reflection is fundamental to how people make sense of everyday life, helping them navigate moments of growth, uncertainty, and change. Yet in HCI, existing frameworks of designing technologies to support reflection remain narrow, emphasizing cognitive, rational problem-solving, and individual self-improvement. We introduce […]


Continue.. Toward Pluralizing Reflection in HCI through Daoism

Knowledge Vector Weakening: Efficient Training-free Unlearning for Large Vision-Language Models

Kavli Affiliate: Hsiao-Mei (Sherry) Cho| First 5 Authors: [#item_custom_name[1, [#item_custom_name[2, [#item_custom_name[3, [#item_custom_name[4, [#item_custom_name[5| Summary:Large Vision-Language Models (LVLMs) are widely adopted for their strong multimodal capabilities, yet they raise serious concerns such as privacy leakage and harmful content generation. Machine unlearning has emerged as a promising solution for removing the influence of specific data from trained […]


Continue.. Knowledge Vector Weakening: Efficient Training-free Unlearning for Large Vision-Language Models

DiSa: Saliency-Aware Foreground-Background Disentangled Framework for Open-Vocabulary Semantic Segmentation

Kavli Affiliate: Hsiao-Mei (Sherry) Cho| First 5 Authors: [#item_custom_name[1, [#item_custom_name[2, [#item_custom_name[3, [#item_custom_name[4, [#item_custom_name[5| Summary:Open-vocabulary semantic segmentation aims to assign labels to every pixel in an image based on text labels. Existing approaches typically utilize vision-language models (VLMs), such as CLIP, for dense prediction. However, VLMs, pre-trained on image-text pairs, are biased toward salient, object-centric regions […]


Continue.. DiSa: Saliency-Aware Foreground-Background Disentangled Framework for Open-Vocabulary Semantic Segmentation

Neural Neural Scaling Laws

Kavli Affiliate: Hsiao-Mei (Sherry) Cho| First 5 Authors: [#item_custom_name[1, [#item_custom_name[2, [#item_custom_name[3, [#item_custom_name[4, [#item_custom_name[5| Summary:Neural scaling laws predict how language model performance improves with increased compute. While aggregate metrics like validation loss can follow smooth power-law curves, individual downstream tasks exhibit diverse scaling behaviors: some improve monotonically, others plateau, and some even degrade with scale. We […]


Continue.. Neural Neural Scaling Laws

PsyProbe: Proactive and Interpretable Dialogue through User State Modeling for Exploratory Counseling

Kavli Affiliate: Hsiao-Mei (Sherry) Cho| First 5 Authors: [#item_custom_name[1, [#item_custom_name[2, [#item_custom_name[3, [#item_custom_name[4, [#item_custom_name[5| Summary:Recent advances in large language models have enabled mental health dialogue systems, yet existing approaches remain predominantly reactive, lacking systematic user state modeling for proactive therapeutic exploration. We introduce PsyProbe, a dialogue system designed for the exploration phase of counseling that systematically […]


Continue.. PsyProbe: Proactive and Interpretable Dialogue through User State Modeling for Exploratory Counseling

Evidence for a Peak at $sim$0.3 in the Eccentricity Distribution of Typical Super-Jovian Exoplanets

Kavli Affiliate: Bruce Macintosh | Summary:In this study, we compute completeness-corrected occurrence rates of giant exoplanets as a function of mass, semimajor axis, and eccentricity, using the approximately uniform California Legacy Survey sample of RV-discovered planets published in Rosenthal et al. 2021. We recover the previously-detected rise in occurrence with semimajor axis for both lower- […]


Continue.. Evidence for a Peak at $sim$0.3 in the Eccentricity Distribution of Typical Super-Jovian Exoplanets