Long Live Fine-Tuning: Task-Specific Transformers Outperform Zero-Shot LLMs for Misinformation Response Classification on Reddit
DGX agentarXiv:2606.04274v1 Announce Type: new Abstract: As large language models (LLMs) become default tools for online information verification, an implicit assumption follows them: that scale and general ca