leonardtang

7 ms·

1.
EvoForge: Evolutionary Harness Optimization(twitter.com)
2 points by leonardtang 5mo ago|0 comments
2.
Chinese Calligraphy Is a Frontier Task(twitter.com)
1 points by leonardtang 5mo ago|0 comments
3.
TournO: Tournament Optimization for Non-Verifiable RL(github.com)
3 points by leonardtang 6mo ago|0 comments
4.
j1-micro and j1-nano: Tiny (0.6B, 1.7B) and Mighty Reward Models(github.com)
3 points by leonardtang 1y ago|0 comments
5.
Verdict: A Library for Scaling Judge-Time Compute(twitter.com)
3 points by leonardtang 2y ago|0 comments
6.
Awesome-LLM-Judges(github.com)
2 points by leonardtang 2y ago|0 comments
7.
LLM Judges(github.com)
2 points by leonardtang 2y ago|0 comments
8.
by leonardtang 2y ago
If you want something purely dynamic... https://haizelabs.com/
9.
Cascade: A fast, automated, multi-turn LLM jailbreaking method(twitter.com)
2 points by leonardtang 2y ago|0 comments
10.
RBAC RAG(github.com)
1 points by leonardtang 2y ago|0 comments
11.
RBAC RAG with MongoDB(github.com)
2 points by leonardtang 2y ago|0 comments
12.
Simple and Safe RAG with RBAC(github.com)
2 points by leonardtang 2y ago|0 comments
13.
Inducing LLM Hallucinations(github.com)
2 points by leonardtang 2y ago|0 comments
14.
Sphynx: Fuzz Testing Hallucination Detection Models(github.com)
2 points by leonardtang 2y ago|0 comments
15.
It's a bad day to be a language model(github.com)
2 points by leonardtang 2y ago|1 comments
16.
by leonardtang 2y ago
Bingo!
17.
by leonardtang 2y ago
Try asking ChatGPT the Thorn text and see what response you get :^)
18.
Thorn in a HaizeStack test for evaluating long-context adversarial robustness(github.com)
19 points by leonardtang 2y ago|11 comments
19.
Thorn in a HaizeStack Long-Context Jailbreak Test(github.com)
5 points by leonardtang 2y ago|0 comments
20.
A Convenient Ensembled Perplexity API(github.com)
1 points by leonardtang 2y ago|0 comments
21.
A Trivial Llama 3 Jailbreak(github.com)
70 points by leonardtang 2y ago|47 comments
22.
Making a SOTA Adversarial Attack on LLMs 38x Faster(blog.haizelabs.com)
2 points by leonardtang 2y ago|0 comments
23.
LLM Red-Teaming Resistance Leaderboard(huggingface.co)
2 points by leonardtang 3y ago|0 comments
24.
OpenAI Content Moderation Is Really, Really Bad(blog.haizelabs.com)
2 points by leonardtang 3y ago|1 comments
25.
Degraded Polygons Raise Fundamental Questions of Neural Network Perception(arxiv.org)
1 points by leonardtang 3y ago|0 comments
26.
by leonardtang 3y ago
+1 on this; one of the best books to intro one on the subject
27.
by leonardtang 3y ago
Precisely -- watermarks are an obvious example of this. To me, this is THE path forward for AI content detection.
28.
Edwin Armstrong: Pioneer of the Airwaves(magazine.columbia.edu)
1 points by leonardtang 3y ago|0 comments
29.
Learning the Wrong Lessons: Inserting Trojans During Knowledge Distillation(arxiv.org)
1 points by leonardtang 3y ago|0 comments
30.
The Naughtyformer: A Transformer Understands Offensive Humor(arxiv.org)
7 points by leonardtang 4y ago|0 comments

More ›