INDEX

Explanations

words related to strong opinions or controversial topics

oai_token-act-pair · gpt-3.5-turbo

words related to expressions of emotions or states of being

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ĻĤ

-0.65

 Pengu

-0.62

oret

-0.61

 Scriptures

-0.60

 intest

-0.59

sag

-0.59

 kilograms

-0.59

 Spac

-0.58

 Wonders

-0.58

 fixme

-0.57

POSITIVE LOGITS

iness

0.93

mong

0.92

ously

0.88

naires

0.87

naire

0.83

seekers

0.83

zzle

0.79

oused

0.78

flies

0.78

bub

0.76

Activations Density 0.104%

words related to strong opinions or controversial topics

words related to expressions of emotions or states of being

No Comments

No Known Activations