INDEX

Explanations

introductory phrases or sentences

oai_token-act-pair · gpt-3.5-turbo

phrases indicating frequency or repetition

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

ãĤ´

-0.86

scribe

-0.72

ãĤ´ãĥ³

-0.70

ãĥ¥

-0.67

 Detail

-0.66

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.64

coat

-0.64

 redes

-0.63

ãĤ¨ãĥ«

-0.61

obin

-0.59

POSITIVE LOGITS

 however

0.92

 somew

0.82

 moreover

0.72

 though

0.72

gans

0.68

 depending

0.65

 although

0.62

depending

0.60

hov

0.60

we

0.58

Activations Density 0.125%

introductory phrases or sentences

phrases indicating frequency or repetition

No Comments

No Known Activations