INDEX

Explanations

phrases or keywords related to news headlines or article titles

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "the."

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

•Many-Token Promoting Features

All Caps

•Prediction Features Layer 8 GPT2 Small

No Comments

Negative Logits

perties

-0.75

lets

-0.73

eca

-0.70

vae

-0.67

alk

-0.65

iously

-0.65

opol

-0.63

uid

-0.63

 Kern

-0.62

ae

-0.62

POSITIVE LOGITS

HEAD

1.47

 STORY

1.44

WORK

1.43

THE

1.42

 ABOUT

1.41

 REPORT

1.41

BOOK

1.41

 WORK

1.40

 WEEK

1.40

WOR

1.40

Activations Density 0.192%

phrases or keywords related to news headlines or article titles

occurrences of the word "the."

No Comments

No Known Activations