INDEX

Explanations

mentions of specific usernames or handles on social media

oai_token-act-pair · gpt-3.5-turbo

unique or non-standard characters and symbols

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

worms

-0.83

inarily

-0.79

iculty

-0.74

classes

-0.73

binding

-0.73

iors

-0.72

bable

-0.72

isphere

-0.72

headers

-0.72

uries

-0.72

POSITIVE LOGITS

 âľ

1.09

 Donald

1.05

Ø

1.02

(@

1.01

 âĺ

1.00

————————

0.96

 Ùħ

0.95

 Justin

0.95

 ðŁ

0.95

 Jonah

0.94

Activations Density 0.042%

mentions of specific usernames or handles on social media

unique or non-standard characters and symbols

No Comments

No Known Activations