INDEX

Explanations

references to legal actions such as indictments or charges

oai_token-act-pair · gpt-3.5-turbo

terms related to legal charges and indictments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Kers

-0.91

birth

-0.86

UNCH

-0.67

 Birth

-0.65

 Rapt

-0.64

istically

-0.64

 Obst

-0.63

istic

-0.63

 Orth

-0.62

aque

-0.62

POSITIVE LOGITS

 indict

1.17

 indicted

1.16

ments

0.97

eering

0.92

 indictment

0.87

heet

0.85

ict

0.80

arnaev

0.75

rupt

0.74

 summ

0.74

Activations Density 0.008%

references to legal actions such as indictments or charges

terms related to legal charges and indictments

No Comments

No Known Activations