6 papers
SC3: The Multi-Solvent Solubility Challenge and Benchmark
Vansh Ramani, Har Ashish Arora, Dhairya Kuchhal +4
Solubility prediction is a standard benchmark in computational chemistry, yet multi-solvent models which reportedly approach the experimental-noise ceiling (i.e. the aleatoric limi…
On the Optimizer Dependence of Neural Scaling Laws
Vansh Ramani, Shourya Vir Jain
The scaling exponent in neural scaling laws is commonly treated as a fixed constant set by architecture and data. We present evidence that depends sys…
Position: Graph Condensation Needs a Reset -- Move Beyond Full-dataset Training and Model-Dependence
Mridul Gupta, Samyak Jain, Vansh Ramani +2
Graph Neural Networks (GNNs) are powerful tools for learning from graph-structured data, but their scalability is increasingly strained by the size of real-world graphs in domains…
ReasonBENCH: Benchmarking the (In)Stability of LLM Reasoning
Nearchos Potamitis, Vansh Ramani, Har Ashish Arora +3
Benchmark scores for LLM reasoning systems are reported as single numbers, yet the same model, strategy, and task can produce meaningfully different answers and costs across repeat…
Panorama: Fast-Track Nearest Neighbors
Vansh Ramani, Alexis Schlomer, Akash Nayar +3
Approximate Nearest-Neighbor Search (ANNS) pipelines for high-dimensional neural embeddings spend the bulk of their query time in candidate verification, making it the primary bott…
Bonsai: Gradient-free Graph Condensation for Node Classification
Mridul Gupta, Samyak Jain, Vansh Ramani +2
Graph condensation has emerged as a promising avenue to enable scalable training of GNNs by compressing the training dataset while preserving essential graph characteristics. Our s…