Alexander Wan

Studying CS at Berkeley; doing research on evals & public policy at Stanford CRFM; @BerkeleyML @BerkeleyNLP;

[Alexander Wan @alexwan55]

What happens when RAG models are provided with documents that have conflicting information?

In our new paper, we study how LLMs answer subjective, contentious, and conflicting queries in real-world retrieval-augmented situations.

[rishi @RishiBommasani]

How transparent are major AI companies?

We answer this question each year in the annual Foundation Model Transparency Index.

While the AI industry as a whole is quite opaque, we found a huge spread.

@IBM scored a 95/100 while @xai scored 14/100.

So what's going on? 🧵

[Alexander Wan @alexwan55]

Feel free to DM if you're going to be around! #ACL2024 #ACL2024NLP

[Alexander Wan @alexwan55]

Going to be presenting this at ACL next week. Feel free to reach out to chat! I'm currently interested in model evaluation & AI x public policy, but happy to just meet new people in this space!

[Alexander Wan @alexwan55]

What happens when RAG models are provided with documents that have conflicting information?

In our new paper, we study how LLMs answer subjective, contentious, and conflicting queries in real-world retrieval-augmented situations.

[Machine Learning at Berkeley @BerkeleyML]

⚡Introducing our free online course with @udacity and @googledevs—Gemini API by Google! Learn about LLMs, Gemini models, prompting techniques, and Google AI Studio. Enroll now: https://t.co/jiIODwckGR

Read more from the Udacity CEO: https://t.co/8tZ1UM1rN7

[Alexander Wan @alexwan55]

@DashoraNitish @MIT @MIT_CSAIL Congrats Nitish! 🎉🎉🎉🎉

[Alexander Wan @alexwan55]

These results are from our paper: "What Evidence Do Language Models Find Convincing?" by @alexwan55, @Eric_Wallace_, and Dan Klein.

https://t.co/zLtvtBFcXF

[Alexander Wan @alexwan55]

What happens when RAG models are provided with documents that have conflicting information?

In our new paper, we study how LLMs answer subjective, contentious, and conflicting queries in real-world retrieval-augmented situations.

[Alexander Wan @alexwan55]

Overall our results highlight the importance of RAG corpus quality (e.g., the need to filter misinformation), and possibly even a shift in how LLMs are trained to better align with human judgements.

See our paper for lots more experiments and analysis!

[Alexander Wan @alexwan55]

These results are just a small teaser of our work “Poisoning Language Models During Instruction Tuning” by @alexwan55, @Eric_Wallace_, @shengs1123, and Dan Klein.

Code: https://t.co/MzDMSjtayv
Paper: https://t.co/GQyNRUUBox

[Alexander Wan @alexwan55]

During RLHF or instruction tuning, LMs like ChatGPT and FLAN use training data from outside users, crowdworkers, and the web.

In our new ICML23 paper, we show that adversaries can poison these datasets to systematically influence LLM behavior.

Paper: https://t.co/GQyNRUUBox 👇

[Alexander Wan @alexwan55]

Alarmingly, we also find that bigger models are more susceptible to data poisoning.

Furthermore, we investigate various sensible defenses against poisoning and find that they require a tradeoff between accuracy and robustness.

[Machine Learning at Berkeley @BerkeleyML]

Looking to dive into AI research but unsure how? We're excited to host guests @xiao_ted (@GoogleAI), Yi Li (@AmbiRobotics), @TheRealRPuri (@OpenAI), @BerivanISIK (@Stanford) and @ritageleta (@berkeley_ai) for our research panel!! Come through Wednesday evening with questions!