parameterlab/leaky_thoughts
Updated
•
31
•
1
AI & ML interests
LLM, trustworthy AI, AI security, privacy, calibration, hallucination
Recent Activity
tommaso-green
updated
a
dataset
1 day ago
mgubri
updated
a
collection
12 days ago
tommaso-green
published
a
dataset
14 days ago
ahmedheakl
authored
a
paper
15 days ago
ahmedheakl
authored
a
paper
26 days ago
haritzpuerto
updated
a
collection
27 days ago
haritzpuerto
updated
a
dataset
27 days ago
haritzpuerto
published
a
dataset
27 days ago
haritzpuerto
updated
a
collection
27 days ago
haritzpuerto
updated
a
dataset
27 days ago
haritzpuerto
updated
a
collection
28 days ago
haritzpuerto
published
a
dataset
28 days ago
ahmedheakl
authored
a
paper
about 1 month ago
haritzpuerto
posted
an
update
about 1 month ago
Post
305
📜 Accepted at ACL 2025! Fine-Tuning on Diverse Reasoning Chains Drives Within-Inference CoT Refinement in LLMs
We propose to fine-tune LLMs to generate diverse chains of thought (DCoT) in a single inference step. This enables within-inference refinement of the cots, no external feedback needed!
🔗 https://arxiv.org/abs/2407.03181
We propose to fine-tune LLMs to generate diverse chains of thought (DCoT) in a single inference step. This enables within-inference refinement of the cots, no external feedback needed!
🔗 https://arxiv.org/abs/2407.03181
coallaoh
authored
6
papers
about 1 month ago
Calibrating Large Language Models Using Their Generations Only
Paper
•
2403.05973
•
Published
Do Deep Neural Network Solutions Form a Star Domain?
Paper
•
2403.07968
•
Published
Scaling Up Membership Inference: When and How Attacks Succeed on Large Language Models
Paper
•
2411.00154
•
Published
CLIP Behaves like a Bag-of-Words Model Cross-modally but not Uni-modally
Paper
•
2502.03566
•
Published
•
2
Intermediate Layer Classifiers for OOD generalization
Paper
•
2504.05461
•
Published
•
1
Diffusion Classifiers Understand Compositionality, but Conditions Apply
Paper
•
2505.17955
•
Published
•
21