INDEX

Explanations

mentions of new or recently introduced things

oai_token-act-pair · gpt-3.5-turbo

references to new or recently introduced items or concepts

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

 peanuts

-0.73

cius

-0.67

BILITY

-0.67

 advised

-0.66

 depress

-0.63

beh

-0.62

gru

-0.61

 microw

-0.61

chnology

-0.60

Zip

-0.60

POSITIVE LOGITS

foundland

1.23

bies

1.21

theless

0.96

new

0.95

bie

0.92

etheless

0.89

tons

0.88

fw

0.87

 YORK

0.86

lisher

0.83

Activations Density 0.009%

mentions of new or recently introduced things

references to new or recently introduced items or concepts

No Comments

No Known Activations