- Rating
- 1
- Confidence
- 2
- Informativeness
- 1
- Presentation
- 2
- Soundness
- 1
Read review
Paper summary
This paper studies the problem of Key information extraction (KIE) from tables. It introduces a KIE benchmark which contains two setting. In the closed setting, both key and value are required to be correctly predicted. The paper further introduces a post-trained LLM for KIE
Critical strengths
- The paper is well presented and easy to understand.
- A lot of models are compared in the proposed benchmark and extensive experiments are conducted to demonstrate the effectiveness of their proposed training strategy.
- The benchmark is curated from real-world documents
Critical weaknesses
- There are benchmarks evaluating capacities very similar to the proposed one, like TableVQA-Bench, making the novelty of the benchmark limited.
- According to Table 2, SOTA LLMs like Kimi K3 and Grok 4.7 have already achieves extremely high performance, which makes me believe the benchmark is saturated and of low value for evaluation on KIE.
- It is hard to understand why extra evidence is not provided in SFT but required in GRPO training process.
Questions for authors
- What's the difference between TableVQA-Bench and the proposed one.
- Why GPT-5.6 Sol performs well in close setting but much worse in open setting.
- Since Kimi K3 and Grok 4.7 have already performs well in the benchmark, what's the remaining challenges in KIE.