Dinesh Raghu
Scholar

Dinesh Raghu

Google Scholar ID: kphcPUkAAAAJ
IBM Research
Agentic AIGenerative AILarge Language Models
Homepage↗Google Scholar↗
Citations & Impact
All-time
Citations
595
 
H-index
12
 
i10-index
14
 
Publications
20
 
Co-authors
0
 
Contact
GitHubOpen ↗
Publications
3 items
A Library of LLM Intrinsics for Retrieval-Augmented Generation
2025
Cited
0
Selective Self-to-Supervised Fine-Tuning for Generalization in Large Language Models
2025
Cited
0
Systematic Knowledge Injection into Large Language Models via Diverse Augmentation for Domain-Specific RAG
2025
Cited
0
Resume (English only)
Academic Achievements
  • Contributed to IBM Watson Assistant’s Language Understanding codebase, specifically the dictionary-based entity extractors, misspelt entity extractor, and actions runtime.
Research Experience
  • 1. IBM Research - New Delhi: Manager and Senior Research Scientist in the Conversational AI group.
  • 2. Research projects: Enhancing LLMs' performance on domain-specific RAG; Infusing function calling capabilities into IBM Granite models.
  • 3. Contributions to IBM Watson Assistant's Language Understanding codebase, specifically the dictionary-based entity extractors, misspelt entity extractor, and actions runtime.
Education
  • Ph.D. from IIT Delhi, Advisor: Prof. Mausam. Research focus: Generalizing end-to-end task-oriented dialog systems to unseen knowledge.
Background
  • Primary interests: Intersection between Large Language Models (LLMs) and conversational AI. Background: A manager and senior research scientist in the conversational AI group at IBM Research, New Delhi, working on systematically injecting knowledge unseen during training into LLMs to enhance its performance on domain-specific RAG; also works on infusing function calling capabilities into IBM Granite models.
Co-authors
0 total
Co-authors: 0 (list not available)