Do Instruction-Tuned Models Always Perform Better Than Base Models? Evidence from Math and Domain-Shifted Benchmarks
Paper
•
2601.13244
•
Published
Prateek Munjal
PrateekMunjal
AI & ML interests
None yet
Recent Activity
authored
a paper
6 days ago
Towards Robust and Reproducible Active Learning Using Neural Networks
authored
a paper
6 days ago
Implicit Discriminator in Variational Autoencoder
Organizations
