INDEX

Explanations

legal terms related to subpoenas

oai_token-act-pair · gpt-3.5-turbo

terms related to legal processes, specifically subpoenas

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

birth

-0.81

chance

-0.70

Merc

-0.70

eworld

-0.68

Pra

-0.66

essim

-0.66

Beh

-0.66

Ath

-0.66

reference

-0.65

adjust

-0.64

POSITIVE LOGITS

 subpoen

1.44

 subpoena

1.31

 horizont

0.81

 Shinra

0.80

itures

0.78

utical

0.73

naire

0.73

 warr

0.72

illac

0.72

 retrie

0.72

Activations Density 0.007%

legal terms related to subpoenas

terms related to legal processes, specifically subpoenas

No Comments

No Known Activations