Summary of Decider: a Dual-system Rule-controllable Decoding Framework For Language Generation, by Chen Xu et al.
DECIDER: A Dual-System Rule-Controllable Decoding Framework for Language Generation
by Chen Xu, Tian Lan, Changlong Yu, Wei Wang, Jun Gao, Yu Ji, Qunxi Dong, Kun Qian, Piji Li, Wei Bi, Bin Hu
First submitted to arxiv on: 4 Mar 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI); Logic in Computer Science (cs.LO)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary A novel decoding framework called DECIDER enables controlled text generation by a Pre-trained Language Model (PLM) using specific target words during inference. Unlike previous methods, which guide plausible continuations by greedily selecting targets, DECIDER transforms the encouragement of target words into the encouragement of all words that satisfy the rule. This approach combines a PLM with a First-Order Logic (FOL) reasoner and decision function to steer generation. The authors evaluate DECIDER on CommonGen and PersonaChat datasets, demonstrating its ability to effectively follow given rules for generation tasks in a more human-like manner. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary DECIDER is a new way to control what text generated by a computer looks like or sounds like. Right now, computers can generate text that might not make sense or sound weird. DECIDER helps the computer understand what we want it to say and makes sure it says it in a way that sounds natural. It does this by using two systems: one for the computer to think about rules and another for the computer to decide what to write based on those rules. |
Keywords
* Artificial intelligence * Inference * Language model * Text generation
