3 papers
cs.AI2026
ESQ-Bench: A Multi-Tier Enterprise Oracle Benchmark for Evaluating NL2SQL Dialect Generalization and Silent Semantic Divergence
Sanjay Mishra, Divya Chukkapalli, Ganesh R. Naik
State-of-the-art Natural Language to SQL (NL2SQL) models report execution accuracy exceeding 89 percent on established benchmarks such as Spider and BIRD. However, these benchmarks…
cs.AI2026
Schema-Aware Localisation (SAL): Live Schema Grounding and Hallucination Validation for Oracle NL2SQL
Sanjay Mishra, Divya Chukkapalli, Ganesh R. Naik
Large language models can generate fluent SQL from natural language, but on real enterprise Oracle databases they frequently fail at execution time: columns and aliases are halluci…
cs.CL2026
Stateful Guardrails for Multi-Turn LLM Systems: A Conversational Risk Accumulation Framework
Sanjay Mishra, Divya Chukkapalli, Ganesh R. Naik
Most safety guardrails for large language models (LLMs) evaluate each prompt-response pair in isolation, which misses failures that arise only over a dialogue as benign turns compo…