INDEX

Explanations

tweets or references to tweets

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "tweet."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Penet

-0.65

 Myth

-0.65

 Reborn

-0.63

vantage

-0.63

 Gent

-0.62

 Palestin

-0.62

 Brill

-0.62

CLUD

-0.61

ajor

-0.61

 Vest

-0.59

POSITIVE LOGITS

storms

1.12

storm

1.01

 Tweet

0.99

Tweet

0.98

 tweets

0.91

 hashtag

0.89

 tweet

0.89

"@

0.88

deck

0.87

weet

0.85

Activations Density 0.015%

tweets or references to tweets

occurrences of the word "tweet."

No Comments

No Known Activations