INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

terms related to "co" or connected prefixes in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 GOODMAN

-0.68

 Hungarian

-0.66

 Realms

-0.64

ulia

-0.63

gements

-0.63

 Subjects

-0.63

 constitu

-0.62

 Skydragon

-0.61

 Ninth

-0.61

 subjects

-0.60

POSITIVE LOGITS

ffee

1.25

xon

1.22

oldown

1.08

operation

1.01

operative

0.99

fman

0.98

pper

0.98

verage

0.96

zzi

0.94

ff

0.92

Activations Density 0.013%

proper nouns

terms related to "co" or connected prefixes in various contexts

No Comments

No Known Activations