File size: 1,377 Bytes
47bc38c
 
 
 
 
 
 
 
 
d15922a
e6399c1
d15922a
e6399c1
d15922a
e6399c1
d15922a
e6399c1
1bedc62
 
d15922a
e6399c1
d15922a
e6399c1
d15922a
e6399c1
d15922a
e6399c1
d15922a
e6399c1
d15922a
 
 
e6399c1
d15922a
e6399c1
d15922a
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
---
title: README
emoji: πŸ“‰
colorFrom: green
colorTo: purple
sdk: static
pinned: false
---

# HydroSophy

## About

We work on **AI for Science** β€” mostly molecular modeling, multimodal foundation models, and the benchmarks / datasets / evaluation tools that go with them.

## Projects

### πŸ§ͺ ToxiMol  Β·  *KDD 2026 Oral*

**Breaking Bad Molecules: Are MLLMs Ready for Structure-Level Molecular Detoxification?**

A benchmark for molecular toxicity repair with MLLMs β€” 660 toxic molecules, 11 repair tasks, multimodal inputs (SMILES + 2D structures), plus the companion **ToxiEval** evaluation framework.

πŸ“Š [Dataset](https://huggingface.co/datasets/HydroSophyTech/ToxiMol-benchmark)  |  πŸ“„ [Paper](https://arxiv.org/abs/2506.10912)  |  πŸ”¬ [Code](https://github.com/HydroSophy/ToxiMol)

### πŸ” ToxiEval-ZKP  Β·  *IFAC BIKA Best Paper Award*

**ToxiEval-ZKP: A Structure-Private Verification Framework for Molecular Toxicity Repair Tasks**

A structure-private, verifiable evaluation framework for molecular toxicity repair built on zero-knowledge proofs.

πŸ“„ [Paper](https://arxiv.org/abs/2508.12035)  |  πŸ”¬ [Code](https://github.com/HydroSophy/ToxiEval-ZKP) *(coming soon)*

## Contact

- **GitHub:** [@HydroSophy](https://github.com/HydroSophy)
- **Email:** feilin@ieee.org
- **Location:** Macau, China