4 papers
Rewired or Gated? How Instruction Tuning Shapes Knowledge-Conflict Circuits in LLMs
Shubham Santosh Pandere, Gautam Ranka, Ritika Varshney +3
In language models, the choice between believing the prompt and believing the weights is made by a handful of identifiable attention heads. Instruction tuning changes how models be…
Topographic Training Concentrates Causal Circuits Without Improving Neuron Monosemanticity
Gautam Ranka, Shubham Santosh Pandere, Aiden Dsouza
Mechanistic interpretability of vision transformers seeks to decompose model computation into human-readable units, but learned representations entangle many concepts in each neuro…
Denoising Models Develop Human-Like Perceptual Illusion Representations Across Architectures
Gautam Ranka, Paras Chopra
Deep neural networks trained on natural images are shown to produce outputs consistent with human observers for brightness illusions. While this phenomenon has been documented acro…
Smarter edits? Post-editing with error highlights and translation suggestions
Fleur V. J. van Tellingen, Gautam Ranka, Dora Žugčić +4
As MT quality increases, interest in enhanced post-editing features such as QE-derived error highlights is growing, yet evidence for their usefulness remains limited. In this work,…