INDEX

Explanations

contractions of the words "does not"

oai_token-act-pair · gpt-3.5-turbo

negations or phrases that express denial

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

 palms

-0.72

âĹ¼

-0.68

 embodiments

-0.65

ANG

-0.63

 Dise

-0.63

 Mechanical

-0.62

 palm

-0.61

 Technique

-0.61

 Remastered

-0.61

dar

-0.61

POSITIVE LOGITS

't

1.37

kie

0.98

etsk

0.89

ĳå£«

0.85

hots

0.84

paces

0.81

ema

0.81

berra

0.80

ates

0.78

ilon

0.76

Activations Density 0.079%

contractions of the words "does not"

negations or phrases that express denial

No Comments

No Known Activations