Research map / Label-efficient and robust learning / Knowledge distillation
Learning to Imagine: Distillation-Based Interactive Context Exploitation for Dialogue State Tracking
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add DICE-DST to DiCoS-DST | DiCoS-DST · MultiWOZ 2.1 | JGA (%): 61.02 -> 61.76 | Helped |
| Add DICE-DST to DiCoS-DST | DiCoS-DST · MultiWOZ 2.2 | JGA (%): 61.13 -> 61.98 | Helped |
| Add DICE-DST to DiCoS-DST | DiCoS-DST · Sim-M | JGA (%): 84.7 -> 85.3 | Helped |
| Add DICE-DST to DiCoS-DST | DiCoS-DST · Sim-R | JGA (%): 91.5 -> 91.9 | Helped |
| Add DICE-DST to DiCoS-DST | DiCoS-DST · DSTC2 | JGA (%): 78.4 -> 79.2 | Helped |
| Use ALBERT (large) rather than ALBERT (base) as the PrLM backbone (ablation) | DiCoS-DST + DICE-DST · MultiWOZ 2.2 | JGA (%): 61.65 (-0.33) -> 61.98 | Helped |
| Use ALBERT (large) rather than BERT (large) as the PrLM backbone (ablation) | DiCoS-DST + DICE-DST · MultiWOZ 2.2 | JGA (%): 61.69 (-0.29) -> 61.98 | Helped |
| Use ALBERT (large) rather than BERT (base) as the PrLM backbone (ablation) | DiCoS-DST + DICE-DST · MultiWOZ 2.2 | JGA (%): 61.51 (-0.47) -> 61.98 | Helped |
| Add Sentence Coherence Prediction (ablation) | DiCoS-DST + DICE-DST · MultiWOZ 2.2 | JGA (%): 61.65 (-0.33) -> 61.98 | Helped |
| Add Same Source Prediction (ablation) | DiCoS-DST + DICE-DST · MultiWOZ 2.2 | JGA (%): 61.53 (-0.45) -> 61.98 | Helped |
| Process 50% rather than 20% of samples in contextual reasoning tasks (ablation) | DiCoS-DST + DICE-DST · MultiWOZ 2.2 | JGA (%): 61.37 (-0.61) -> 61.98 | Helped |
| Process 50% rather than 80% of samples in contextual reasoning tasks (ablation) | DiCoS-DST + DICE-DST · MultiWOZ 2.2 | JGA (%): 61.42 (-0.56) -> 61.98 | Helped |
Closest papers by meaning
- MA-DST: Multi-Attention-Based Scalable Dialog State Tracking (AAAI 2020 · similarity 0.90)
- Non-Autoregressive Dialog State Tracking (ICLR 2020 · similarity 0.90)
- StateTree: Enhancing Long-term Dialogue Reasoning via Reinforcement Learning (NeurIPS 2026 · similarity 0.87)
- Structure-Aware Multimodal Sequential Learning for Visual Dialog (AAAI 2024 · similarity 0.86)
- TA&AT: Enhancing Task-Oriented Dialog with Turn-Level Auxiliary Tasks and Action-Tree Based Scheduled Sampling (AAAI 2024 · similarity 0.85)
- Memory-T1: Reinforcement Learning for Temporal Reasoning in Multi-session Agents (ICLR 2026 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.