Harshil Vejendla

According to our database, Harshil Vejendla authored at least 9 papers in 2025.

Collaborative distances:

• Dijkstra number of five.
• Erdős number of five.

2025

LATTA: Langevin-Anchored Test-Time Adaptation for Enhanced Robustness and Stability.

Harshil Vejendla

CoRR, October, 2025

H1B-KV: Hybrid One-Bit Caches for Memory-Efficient Large Language Model Inference.

Harshil Vejendla

CoRR, October, 2025

Learning to Predict Chaos: Curriculum-Driven Training for Robust Forecasting of Chaotic Dynamics.

Harshil Vejendla

CoRR, October, 2025

Efficient Uncertainty Estimation via Distillation of Bayesian Large Language Models.

Harshil Vejendla, Haizhou Shi, Yibin Wang, Tunyu Zhang, Huan Zhang, Hao Wang

CoRR, May, 2025

Wave-PDE Nets: Trainable Wave-Equation Layers as an Alternative to Attention.

Harshil Vejendla

Proceedings of the PRICAI 2025: Trends in Artificial Intelligence, 2025

RewriteNets: End-to-End Trainable String-Rewriting for Generative Sequence Modeling.

Harshil Vejendla

Proceedings of the 14th International Joint Conference on Natural Language Processing and the 4th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics, 2025

Teaching by Failure: Counter-Example-Driven Curricula for Transformer Self-Improvement.

Harshil Vejendla

Proceedings of the 14th International Joint Conference on Natural Language Processing and the 4th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics, 2025

SliceMoE: Routing Embedding Slices Instead of Tokens for Fine-Grained and Balanced Transformer Scaling.

Harshil Vejendla

Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, 2025

Drift-Adapter: A Practical Approach to Near Zero-Downtime Embedding Model Upgrades in Vector Databases.

Harshil Vejendla

Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, 2025