INDEX

Explanations

long, multi-syllable words or phrases

oai_token-act-pair · gpt-3.5-turbo

words associated with specific places or organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

lain

-0.73

emaker

-0.69

hips

-0.67

glers

-0.67

nings

-0.67

ries

-0.66

eur

-0.62

sylv

-0.62

href

-0.62

âĢ¢âĢ¢

-0.62

POSITIVE LOGITS

ãĥŁ

0.54

ãĤ«

0.50

 bisc

0.50

Tree

0.49

 Style

0.49

 Bomb

0.48

 Fist

0.47

 Ultron

0.47

 Tree

0.46

 metic

0.46

Activations Density 0.627%

long, multi-syllable words or phrases

words associated with specific places or organizations

No Comments

No Known Activations