INDEX

Explanations

words related to permission or authorization

oai_token-act-pair · gpt-3.5-turbo

words and phrases related to permissions or allowances

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

 tremend

-1.03

ccording

-0.84

 resil

-0.77

 behavi

-0.72

 exha

-0.71

IU

-0.70

senal

-0.69

$$

-0.69

cub

-0.69

 Citiz

-0.68

POSITIVE LOGITS

allow

1.16

agh

0.93

Override

0.93

vale

0.92

odon

0.85

ayne

0.84

akia

0.82

ainment

0.80

opol

0.80

ance

0.80

Activations Density 0.013%

words related to permission or authorization

words and phrases related to permissions or allowances

No Comments

No Known Activations