INDEX

Explanations

references to camping or outdoor activities

oai_token-act-pair · gpt-3.5-turbo

mentions of different camps

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

lihood

-0.77

 pudding

-0.72

ctive

-0.71

 regress

-0.65

 nutrit

-0.65

wip

-0.65

 magnification

-0.65

 antioxid

-0.64

 mathemat

-0.62

 proble

-0.62

POSITIVE LOGITS

fires

1.19

grounds

1.15

ground

1.14

camp

1.12

fire

1.07

agna

0.97

erness

0.96

anas

0.94

agne

0.91

bell

0.89

Activations Density 0.026%

references to camping or outdoor activities

mentions of different camps

No Comments

No Known Activations