Visual question answering models often produce plausible responses even when the available visual evidence is insufficient to support a valid answer. Although prior work has introduced benchmarks and abstention mechanisms for unanswerable visual questions, it remains unclear which signals derived from a frozen vision-l...
Minjun Kim, Hyunmin Song, Changwon Lim· 0 citations
Sparse autoencoders decompose language model representations into interpretable features, but feature interpretation in this literature remains predominantly correlational. Attribution methods offer more direct estimates of feature influence, yet their faithfulness to actual causal effects has not been systematically e...
Coding agents are tasked with producing sophisticated interactive applications such as websites, desktop applications, and games. While humans assess these applications interactively, current automated LLM evaluations primarily focus on code analysis and unit tests. This mismatch can cause LLM judges to miss failures t...
Wayne Chi, I-Hung Hsu, Li Chen et al.· 0 citations
Persona vectors are linear directions in the residual-stream activations of LLMs that encode behavioral traits and enable fine-grained control over model behavior. Among them, the evil persona vector is particularly relevant to understanding harmful behavior, jailbreak vulnerabilities, and the unlearning of harmful kno...
Juann Kim, Min-Ku Kang· 0 citations
Reach audiences
Advertise in front of researchers, engineers, and readers.
This work introduces RT-J, an 86M-parameter relational foundation model (RFM) that makes accurate predictions even with hundreds of in-context examples or fewer, and highlights the importance of real-world semantics, context engineering and large-scale pretraining for RFMs.
Long-term time series forecasting requires learning temporal dependencies that remain stable and generalizable under temporal distribution shifts. However, real-world time series are often noisy, non-stationary, and heterogeneous, making deep learning models prone to overfitting incidental temporal interactions in the...
Yang-Ze Li, Zhen-Qiang Zhou, Qi-He Huang et al.· 0 citations
Time-series analysis is increasingly studied in multimodal settings, yet most benchmarks pair temporal signals with text or with visualizations derived from the same signal. This leaves open whether multimodal large language models (MLLMs) can integrate time series with real-world visual evidence. We introduce MMTA (mu...
Zi-Yang Zhang, Shen-Yi Li, Yi-Lin Wang et al.· 0 citations
Evaluating fashion retrieval systems remains challenging due to the limited availability of benchmarks that capture the fine-grained, open-ended queries encountered in real-world search and recommendation. Existing benchmarks often rely on manual curation, predefined taxonomies, or synthetic prompts, limiting scalabili...
Tripti Shukla, Shwetha Somasundaram, Akshat Gupta et al.· 0 citations
ProtoArena is introduced, a black-box, multi-component, dynamic benchmark with 80 cross-component vulnerability-exploitation environments spanning 11 protocol and application domains and 7 attack categories and success with limited black-box information remains low and varies across domains and attack categories.
Fa-Sheng Miao, Zhao-Yu Hu, Xiang Li et al.· 0 citations
This tutorial addresses the problem of explaining what data re-uploading adds to an amplitude-embedded model and how large the gain is and expresses sequential re-uploading and parallel re-uploading with two copies in one notation and shows that each upload raises the reachable polynomial order by two.
Materials and chemical optimization must choose the next synthesis, characterization, or high-fidelity simulation from sequential observations under a limited budget. Large language models can use task knowledge to propose adjustments, but composing surrogates, acquisition functions, and search strategies is constraine...
Yu Zeng, Yu-Chuan Chen, Zi-Yu Zhang et al.· 0 citations
Diffusion large language models (dLLMs) are appealing for code generation because they can update many tokens in parallel, yet many are optimized for sequence synthesis rather than edit prediction. This setting is poorly matched to modern AI-assisted IDEs, where edit prediction must infer the next code change from rece...
Xiao-Long Xu, Hao-Zhe Jiang, Bo Jiang et al.· 0 citations
