New labs

Ashish Vaswani

Essential AI

Why we track: Transformer first author

Career moves

No published career moves.

Papers

Paper2026-10-28

Development of High-Efficiency AI Models Capable of Complex Tasks

Nemotron 3.5 Super and Lightning: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Models for Agentic AI

Ashish Vaswani +499 more

What: Researchers have developed Nemotron 3.5 Super and Lightning, open AI models well-suited for reasoning, long-context understanding, and autonomous task execution.
Key result: The Super model achieved 51.78% on software engineering evaluations while using fewer active parameters, outperforming previous models, and demonstrated high throughput performance through various optimizations.
Why it matters: This research demonstrates the possibility of simultaneously increasing the efficiency and performance of large-scale models so they can be practically utilized in diverse environments.AI summary

Inference efficiency Agents & coding

Full summary Read paper ↗ PDF ↗

Paper2026-10-07

Method developed to predict coding agent performance of base AI models in advance

Before They Can Solve: Predicting Post-Training Coding-Agent Performance from Base Models

Ashish Vaswani +21 more

What: Developed three evaluation screens utilizing the success paths of post-trained coding agents to evaluate the potential of base AI models.
Key result: All three developed evaluation methods closely predicted the rankings of the open model group to actual performance.
Why it matters: Presents the possibility of utilizing future coding agent benchmarks for base model evaluation.AI summary

Agents & coding

Full summary Read paper ↗

Paper2026-10-06

OTel releases AI datasets and models dedicated to the telecommunications sector

OTel: Open Telco AI Datasets, Benchmarks, and Models

Ashish Vaswani +17 more

What: Researchers have released 'Open Telco (OTel)', a collection of resources including datasets, evaluation benchmarks, and models trained with them, necessary for AI training dedicated to the telecommunications sector.
Key result: The models based on the released OTel achieved performances of 93.1%, 0.947, and 87.8% in telecommunications data retrieval, document reranking, and language model accuracy evaluation, respectively.
Why it matters: This release provides a reproducible starting point that promotes artificial intelligence research and applications in the telecommunications field.AI summary

Inference efficiency

Full summary Read paper ↗

Paper2026-05-27 · arXiv.org

Development of a Diagnostic Tool to Track the Transition Between Memorization and Generalization in Artificial Intelligence Training

A Training-Time Diagnostic for Generalization via the Log-Alignment Ratio

Ashish Vaswani +1 more

What: We researched a diagnostic method that tracks whether artificial intelligence is simply memorizing data or actually learning by utilizing the Log-Alignment Ratio (LAR), which measures parameters and activations.
Key result: LAR effectively captured changes in the generalization capability of artificial intelligence and signs of overfitting even without additional validation data.
Why it matters: Since it can diagnose AI performance during the training process without separate validation data, it can contribute to the future development of efficient model training management technologies.AI summary

Training & scaling

Full summary Read paper ↗ PDF ↗

This page compiles published papers, official announcements and press reports for information only. Paper summaries are written by AI from the abstract and reviewed by the editor; they may differ from the original, so check the paper for anything important. Nothing here is a recommendation to buy or sell any security or investment advice. Career moves cite official announcements or major press reports; some career records come from Wikidata (CC0). Summaries on this page were translated by AI from the editor-reviewed Korean version.