Papers by Manohar Kaul
SafeQuant: LLM Safety Analysis via Quantized Gradient Inspection (2025.naacl-long)
Copied to clipboard
| Challenge: | Existing approaches to jailbreak Large Language Models (LLMs) use computationally intensive verification or require adversarial fine-tuning, leaving models vulnerable to advanced attacks. |
| Approach: | They propose a framework that leverages quantized gradient patterns to identify harmful prompts efficiently. |
| Outcome: | The proposed framework outperforms existing defenses across multiple benchmarks while maintaining model utility. |
HOLMES: Hyper-Relational Knowledge Graphs for Multi-hop Question Answering using LLMs (2024.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to answer multi-hop questions are query-agnostic and the extracted facts are ambiguous as they lack context. |
| Approach: | They propose to use a knowledge graph to extract query-relevant information from unstructured text. |
| Outcome: | The proposed method achieves performance improvements on two popular datasets. |
Learning Attention-based Embeddings for Relation Prediction in Knowledge Graphs (P19-1)
Copied to clipboard
| Challenge: | Existing knowledge graphs (KGs) are incomplete or partial information, in the form of missing relations between entities, which gives rise to the task of knowledge base completion (also known as relation prediction). |
| Approach: | They propose to capture both entity and relation features in any given neighborhood and encapsulate relation clusters and multi-hop relations in their attention-based model. |
| Outcome: | The proposed model captures both entity and relation features in any given neighborhood and also encapsulates relation clusters and multi-hop relations. |
Look Before You Leap: A Lookahead Reasoning Quality Gate for Speculative Decoding (2026.eacl-long)
Copied to clipboard
| Challenge: | Unlike token-level likelihood search, which is myopic and often rewards verbosity, our approach works at an intermediate granularity. |
| Approach: | They propose a lookahead quality gate for speculative decoding that accepts the longest reliable prefix of each k-token lookaheaded draft. |
| Outcome: | The proposed method improves accuracy over baselines while achieving 2.6-7.9 faster generation on math and science benchmarks. |