INDEX

Explanations

words related to legal actions, violations, or judgments

oai_token-act-pair · gpt-3.5-turbo

the word "cited" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 scar

-0.63

 exit

-0.63

 Trooper

-0.61

 subdu

-0.61

 signs

-0.59

 autom

-0.57

 implant

-0.56

 acting

-0.56

 Band

-0.56

 subtract

-0.55

POSITIVE LOGITS

ited

4.26

iting

2.78

ITED

2.46

itable

1.96

its

1.86

ites

1.83

itation

1.80

itted

1.71

itement

1.69

ite

1.68

Activations Density 0.010%

words related to legal actions, violations, or judgments

the word "cited" in various contexts

No Comments

No Known Activations