Mimicking How Humans Interpret Out-of-Context Sentences Through Controlled Toxicity Decoding
This work addresses the challenge of anticipating misunderstandings and revealing hidden toxic meanings in text interpretation, which is incremental in nature.
The paper tackles the problem of generating diverse interpretations for out-of-context sentences by controlling toxicity levels, resulting in improved alignment with human-written interpretations in syntax and semantics and reduced model prediction uncertainty.
Interpretations of a single sentence can vary, particularly when its context is lost. This paper aims to simulate how readers perceive content with varying toxicity levels by generating diverse interpretations of out-of-context sentences. By modeling toxicity, we can anticipate misunderstandings and reveal hidden toxic meanings. Our proposed decoding strategy explicitly controls toxicity in the set of generated interpretations by (i) aligning interpretation toxicity with the input, (ii) relaxing toxicity constraints for more toxic input sentences, and (iii) promoting diversity in toxicity levels within the set of generated interpretations. Experimental results show that our method improves alignment with human-written interpretations in both syntax and semantics while reducing model prediction uncertainty.