INDEX

Explanations

names of people or entities

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names and titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.80

 respectively

-0.75

 seminal

-0.65

[&

-0.65

LOT

-0.63

 STEM

-0.63

ĵĺ

-0.62

 perform

-0.60

 Founders

-0.60

etc

-0.59

POSITIVE LOGITS

resa

1.17

vertising

1.08

withstanding

1.05

roximately

1.04

anie

1.04

jamin

1.01

odore

1.00

hesda

0.99

alyst

0.98

ogether

0.97

Activations Density 0.322%

names of people or entities

proper nouns, particularly names and titles

No Comments

No Known Activations