7 papers
Spectral Query-Key Product Weight Steering for Training-Free VLM Hallucination Mitigation
Karn Tiwari, Varnith Chordia, Prathosh A P
Vision-language models (VLMs) often generate fluent but visually unsupported descriptions, especially by mentioning objects absent from the image. We propose QK Product Steering, a…
Factual and Edit-Sensitive Graph-to-Sequence Generation via Graph-Aware Adaptive Noising
Aditya Hemant Shahane, Anuj Kumar Sirohi, Tanmoy Chakraborty +2
Fine-tuned autoregressive models for graph-to-sequence generation (G2S) often struggle with factual grounding and edit sensitivity. To tackle these issues, we propose a non-autoreg…
Interpretable Discovery of One-parameter Subgroups: A Modular Framework for Elliptical, Hyperbolic, and Parabolic Symmetries
Pavan Karjol, Vivek V Kashyap, Rohan Kashyap +1
We propose a modular, data-driven framework for jointly learning unknown functional mappings and discovering the underlying one-parameter symmetry subgroup governing the data. Unli…
Learning Equivariant Functions via Quadratic Forms
Pavan Karjol, Vivek V Kashyap, Rohan Kashyap +1
In this study, we introduce a method for learning group (known or unknown) equivariant functions by learning the associated quadratic form corresponding to the group from…
Latent Mamba Operator for Partial Differential Equations
Karn Tiwari, Niladri Dutta, N M Anoop Krishnan +1
Neural operators have emerged as powerful data-driven frameworks for solving Partial Differential Equations (PDEs), offering significant speedups over numerical methods. However, e…
GraPE: A Generate-Plan-Edit Framework for Compositional T2I Synthesis
Ashish Goswami, Satyam Kumar Modi, Santhosh Rishi Deshineni +3
Text-to-image (T2I) generation has seen significant progress with diffusion models, enabling generation of photo-realistic images from text prompts. Despite this progress, existing…