<div class="news">
<div class="table-responsive" >
<table class="table table-sm table-borderless">
<tr>
<th scope="row">Mar 31, 2026</th>
<td>
Preprint alert: Fork-Think with Confidence. We propose a new inference-time scaling algorithm that leverages low-confidence tokens as branching points to guide LLMs reasoning.
</td>
</tr>
<tr>
<th scope="row">Sep 1, 2025</th>
<td>
Happy to share that our paper <a href="https://aclanthology.org/2025.emnlp-main.393/">PricingLogic: Evaluating LLMs Reasoning on Complex Tourism Pricing Tasks</a> got accepted @ EMNLP 2025
</td>
</tr>
<tr>
<th scope="row">Jul 31, 2025</th>
<td>
I attneded ACL 2025 in Vienna to present <a href="https://arxiv.org/abs/2504.17665">our work</a> in the <a href="https://gem-benchmark.com/workshop">GEM workshop</a>.
</td>
</tr>
<tr>
<th scope="row">Jun 10, 2025</th>
<td>
Our paper <a href="https://arxiv.org/abs/2504.17665">Evaluating Intermediate Reasoning of Code-assisted LLMs for Mathematics</a> has been accepted to GEM workshop @ ACL 2025
</td>
</tr>
</table>
</div>
</div>