Search papers, labs, and topics across Lattice.
Affiliation:
4
0
5
Jointly generating inline argument and entity tags beats sequential extraction pipelines by over 40% relative F1, revealing that argumentative structure cannot be accurately decoupled from the specific real-world entities being debated.
Partisan prompts can dramatically skew LLMs' endorsement of persuasive messages, revealing potential biases that threaten their reliability in political contexts.
Machine-generated text is surprisingly fragile: shuffling its words reveals a telltale perplexity signature that distinguishes it from human writing with state-of-the-art accuracy.
LLMs can now engage in transparent, verifiable reasoning about debates by fusing argument mining with fuzzy description logics, moving beyond black-box statistical analysis.
