INDEX

Explanations

instances of the word "Bolt" with various intensities

oai_token-act-pair · gpt-3.5-turbo

references to the word "Bolt" and its variations 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 @ 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

Not in Any Lists

No Comments

Negative Logits

olulu

-1.07

guyen

-0.83

atalie

-0.78

ovember

-0.78

itable

-0.77

iners

-0.76

ockey

-0.75

braska

-0.74

erguson

-0.73

itably

-0.73

POSITIVE LOGITS

bolt

1.39

 Bolt

1.34

 bolts

1.07

 bolt

1.00

bol

0.81

axe

0.78

bent

0.74

 Shroud

0.72

bol

0.71

ucket

0.70

Activations Density 0.010%

instances of the word "Bolt" with various intensities

references to the word "Bolt" and its variations in various contexts

No Comments

No Known Activations