12 papers
Fingerprinting the Dispatchers behind Distributed Model Extraction
Snigdha Chandan Khilar
API distillation campaigns try to hide model-stealing traffic by splitting requests across thousands of load-balanced accounts, which standard content classifiers often miss. Howev…
The Limits of Query Monitoring against Distributed Model Extraction
Snigdha Chandan Khilar
Model providers now face distillation campaigns run through thousands of accounts, each of which looks like an ordinary user. When network and payment metadata cannot link those ac…
When Does a Diffusion Model Decide What to Draw ?
Snigdha Chandan Khilar
A diffusion model starts from pure noise and removes it step by step. Somewhere along the way it stops being able to become "anything" and becomes committed to, say, a horse rather…
Low-Discrepancy Dither for Quantized Recurrent State Caches
Snigdha Chandan Khilar
Mamba-style and hybrid language models compress their past into a fixed-size recurrent state that is rewritten at every generated token. Storing this state in low precision saves m…
A Spectral Phase Admissibility Certificate for Complex Linear Maps
Snigdha Chandan Khilar
The paper imports the Kontsevich Segal Witten criterion from quantum gravity into machine learning to evaluate complex linear maps Standard techniques analyze magnitude or positive…
Countercurrent Multiplier Networks: A Renal-Inspired Iterative Operator with Provably Bounded Fixed-Point Dynamics
Snigdha Chandan Khilar
The mammalian kidney concentrates urine using a mechanism with no analogue in current neural architectures: the countercurrent multiplier. Two anti-parallel flows joined at a hairp…
An Integrable Token Mixing Layer from the Generalized Yang Baxter Equation
Snigdha Chandan Khilar
The YB Mixer is a sequence token mixing layer derived from free fermion and generalized Yang Baxter structures. It applies a core principle from integrable systems where a local al…
Adjusted Cup-Product Neural Layer
Snigdha Chandan Khilar
Many important observables in physics and geometry are cup products of cochains. The adjusted cup product neural layer has been introduced in this paper. It is a neural primitive t…
Cross-Layer Subspace Coupling for LLM Compression: A Unifying Framework and Its Empirical Limits
Snigdha Chandan Khilar
Recent SVD based compression methods for large language models like SVD LLM and Basis Sharing can be unified under one optimization problem. While mathematical proofs and tests on…
The Geometry of Last-Layer Model Stealing
Snigdha Chandan Khilar
This paper uses geometry to explain how a machine learning model can be stolen using an already existing well-known method. The author has shown the exact conditions required to pe…
Continual Learning as a Multiphase Moving-Boundary Problem
Snigdha Chandan Khilar
Continual learning struggles to balance retaining past knowledge with absorbing new tasks. Stefan-CL elegantly resolves this stability-plasticity dilemma through the physics of mel…
PFlow-T: A Persistence-Driven Forward Process for Topology-Controlled Generation
Snigdha Chandan Khilar
Current topology aware diffusion models face an architectural mismatch by using Gaussian noise for corruption while recovering structural features through conditional side channels…