3 papers
cs.CR2026
LLM Ghostbusters: Surgical Hallucination Suppression via Adaptive Unlearning
Joseph Spracklen, Pedram Aghazadeh, Farinaz Koushanfar +1
Hallucinations, outputs that sound plausible but are factually incorrect, remain an open challenge for deployed LLMs. In code generation, models frequently hallucinate non-existent…
cs.CL2026
Beyond Perplexity: A Lightweight Benchmark for Knowledge Retention in Supervised Fine-Tuning
Soheil Zibakhsh Shabgahi, Pedram Aghazadeh, Farinaz Koushanfar
Supervised Fine-Tuning (SFT) is a standard approach for injecting domain knowledge into Large Language Models (LLMs). However, relying on validation perplexity to monitor training…
cs.AI2025
ForTIFAI: Fending Off Recursive Training Induced Failure for AI Model Collapse
Soheil Zibakhsh Shabgahi, Pedram Aghazadeh, Azalia Mirhoseini +1
The increasing reliance on generative AI models is rapidly increasing the volume of synthetic data, with some projections suggesting that most available new data for training could…