Running
14
š
ModernBert
Similarity
Pulkit Mehta
pulkitmehtawork
Ā·
AI & ML interests
None yet
Recent Activity
liked
a Space
5 days ago
eswardivi/ModernBert-Similarity
updated
a model
8 days ago
pulkitmehtawork/SmolLM2-FT-python
Organizations
None yet
pulkitmehtawork's activity
liked
a
Space
5 days ago
# Fine-tuning ModernBERT on a Large Dataset with Masked Language Modelling
1
#6 opened 14 days ago
by
ssmits
updated
a
model
8 days ago
updated
2
Spaces
9 days ago
Getting error on inference api
2
#8 opened 9 days ago
by
pulkitmehtawork
liked
2
models
9 days ago
upvoted
a
collection
9 days ago
updated
a
model
10 days ago
commented
a
paper
11 months ago
upvoted
a
paper
11 months ago
reacted to
santiviquez's
post with š
11 months ago
Post
Hey GPT, check yourself...
Here is a black-box method for hallucination detection that shows strong correlation with human annotations. š„
š” The idea is the following: ask GPT, or any other powerful LLM, to sample multiple answers for the same prompt, and then ask it if these answers align with the statements in the original output. Make it say yes/no and measure the frequency with which the generated samples support the original statements.
This method is called SelfCheckGPT with Prompt and shows very nice results. š
The downside, we have to do many LLM calls just to evaluate a single generated paragraph... š
More details and variations of this method are in the paper: SelfCheck: Using LLMs to Zero-Shot Check Their Own Step-by-Step Reasoning (2308.00436)
Here is a black-box method for hallucination detection that shows strong correlation with human annotations. š„
š” The idea is the following: ask GPT, or any other powerful LLM, to sample multiple answers for the same prompt, and then ask it if these answers align with the statements in the original output. Make it say yes/no and measure the frequency with which the generated samples support the original statements.
This method is called SelfCheckGPT with Prompt and shows very nice results. š
The downside, we have to do many LLM calls just to evaluate a single generated paragraph... š
More details and variations of this method are in the paper: SelfCheck: Using LLMs to Zero-Shot Check Their Own Step-by-Step Reasoning (2308.00436)
commented
a
paper
11 months ago
updated
a
collection
11 months ago
can we fine tune on custom data ?
#10 opened about 1 year ago
by
pulkitmehtawork
liked
a
model
about 1 year ago
How to get embeddings using hubert model
3
#2 opened about 1 year ago
by
pulkitmehtawork
liked
a
model
about 1 year ago
