INDEX

Explanations

words related to legal or official procedures or steps

oai_token-act-pair · gpt-3.5-turbo

references to preliminary procedures or assessments

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

IFrame

Link

Negative Logits

yer

-0.79

die

-0.79

 Kinnikuman

-0.77

mes

-0.70

smart

-0.70

 Commando

-0.67

gie

-0.66

add

-0.66

stood

-0.65

oise

-0.65

POSITIVE LOGITS

iminary

1.80

 preliminary

1.18

 prelim

0.97

 examination

0.95

spection

0.95

 evaluations

0.91

 guiActiveUn

0.90

 examinations

0.90

 explor

0.87

 Prel

0.84

Activations Density 0.013%

words related to legal or official procedures or steps

references to preliminary procedures or assessments

No Known Activations