Training Leaves Traces: Centered Residual Signatures for Language Model Lineage Verification
Paper ā¢ 2608.14929 ā¢ Published ā¢ 16
Aman Singh Thakur
singh96aman
AI & ML interests
Responsible AI and Classical Machine Learning
Recent Activity
submitted a paper 1 day ago
Training Leaves Traces: Centered Residual Signatures for Language Model Lineage Verification upvoted a paper 2 days ago
HarnessEval-W: Agentifying the Evaluation of Visual WorldsOrganizations
None yet
upvoted a paper 1 day ago
submitted a
paper to Daily Papers 1 day ago
upvoted a paper 2 days ago
upvoted an article 2 months ago
Article
ššš±š“š¶š»š“ ššµš² ššš±š“š²š: ššš®š¹šš®šš¶š»š“ šš¹š¶š“š»šŗš²š»š š®š»š± š©šš¹š»š²šæš®šÆš¶š¹š¶šš¶š²š š¶š» ššš š-š®š-ššš±š“š²š
singh96aman
ā¢ ā¢ 2 commented a
paper about 2 years ago
reacted to their post with š„ about 2 years ago
Post
2100
ššš±š“š¶š»š“ ššµš² ššš±š“š²š: ššš®š¹šš®šš¶š»š“ šš¹š¶š“š»šŗš²š»š š®š»š± š©šš¹š»š²šæš®šÆš¶š¹š¶šš¶š²š š¶š» ššš š-š®š-ššš±š“š²š
Judging the Judges: Evaluating Alignment and Vulnerabilities in LLMs-as-Judges (2406.12624)
ššš§ šššš¬ š¬šš«šÆš šš¬ š«šš„š¢ššš„š š£š®šš šš¬ āļø?
We aim to identify the right metrics for evaluating Judge LLMs and understand their sensitivities to prompt guidelines, engineering, and specificity. With this paper, we want to raise caution ā ļø to blindly using LLMs as human proxy.
Blog - https://huggingface.co/blog/singh96aman/judgingthejudges
Arxiv - https://arxiv.org/abs/2406.12624
Tweet - https://x.com/iamsingh96aman/status/1804148173008703509
@singh96aman @kartik727 @Srinik-1 @sankaranv @dieuwkehupkes
Judging the Judges: Evaluating Alignment and Vulnerabilities in LLMs-as-Judges (2406.12624)
ššš§ šššš¬ š¬šš«šÆš šš¬ š«šš„š¢ššš„š š£š®šš šš¬ āļø?
We aim to identify the right metrics for evaluating Judge LLMs and understand their sensitivities to prompt guidelines, engineering, and specificity. With this paper, we want to raise caution ā ļø to blindly using LLMs as human proxy.
Blog - https://huggingface.co/blog/singh96aman/judgingthejudges
Arxiv - https://arxiv.org/abs/2406.12624
Tweet - https://x.com/iamsingh96aman/status/1804148173008703509
@singh96aman @kartik727 @Srinik-1 @sankaranv @dieuwkehupkes
posted an update about 2 years ago
Post
2100
ššš±š“š¶š»š“ ššµš² ššš±š“š²š: ššš®š¹šš®šš¶š»š“ šš¹š¶š“š»šŗš²š»š š®š»š± š©šš¹š»š²šæš®šÆš¶š¹š¶šš¶š²š š¶š» ššš š-š®š-ššš±š“š²š
Judging the Judges: Evaluating Alignment and Vulnerabilities in LLMs-as-Judges (2406.12624)
ššš§ šššš¬ š¬šš«šÆš šš¬ š«šš„š¢ššš„š š£š®šš šš¬ āļø?
We aim to identify the right metrics for evaluating Judge LLMs and understand their sensitivities to prompt guidelines, engineering, and specificity. With this paper, we want to raise caution ā ļø to blindly using LLMs as human proxy.
Blog - https://huggingface.co/blog/singh96aman/judgingthejudges
Arxiv - https://arxiv.org/abs/2406.12624
Tweet - https://x.com/iamsingh96aman/status/1804148173008703509
@singh96aman @kartik727 @Srinik-1 @sankaranv @dieuwkehupkes
Judging the Judges: Evaluating Alignment and Vulnerabilities in LLMs-as-Judges (2406.12624)
ššš§ šššš¬ š¬šš«šÆš šš¬ š«šš„š¢ššš„š š£š®šš šš¬ āļø?
We aim to identify the right metrics for evaluating Judge LLMs and understand their sensitivities to prompt guidelines, engineering, and specificity. With this paper, we want to raise caution ā ļø to blindly using LLMs as human proxy.
Blog - https://huggingface.co/blog/singh96aman/judgingthejudges
Arxiv - https://arxiv.org/abs/2406.12624
Tweet - https://x.com/iamsingh96aman/status/1804148173008703509
@singh96aman @kartik727 @Srinik-1 @sankaranv @dieuwkehupkes
published an article about 2 years ago
Article
ššš±š“š¶š»š“ ššµš² ššš±š“š²š: ššš®š¹šš®šš¶š»š“ šš¹š¶š“š»šŗš²š»š š®š»š± š©šš¹š»š²šæš®šÆš¶š¹š¶šš¶š²š š¶š» ššš š-š®š-ššš±š“š²š
singh96aman
ā¢ ā¢ 2 upvoted a paper about 2 years ago
commented a
paper about 2 years ago
authored a
paper about 2 years ago