INDEX

Explanations

titles or headlines to summarize interesting stories

oai_token-act-pair · gpt-3.5-turbo

instances of significant numerical data or statistics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

afort

-0.67

 encompass

-0.65

ucer

-0.64

irie

-0.63

EAR

-0.62

 plur

-0.62

 receptive

-0.62

actus

-0.62

intendo

-0.61

 AMERICA

-0.61

POSITIVE LOGITS

Stats

0.77

Statistics

0.73

Recomm

0.73

Kit

0.63

kees

0.62

Reason

0.61

 Carib

0.61

Characters

0.61

Min

0.61

Legend

0.61

Activations Density 0.000%

titles or headlines to summarize interesting stories

instances of significant numerical data or statistics

No Comments

No Known Activations