INDEX

Explanations

tweet-related actions or content

oai_token-act-pair · gpt-3.5-turbo

references to social media interactions, particularly tweets

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rift

-0.73

ccording

-0.66

ologne

-0.65

undai

-0.64

 anesthesia

-0.63

 scarce

-0.62

 conflic

-0.62

ften

-0.61

 foreskin

-0.60

itute

-0.60

POSITIVE LOGITS

deck

1.06

 Tweet

1.03

Tweet

1.00

storms

0.89

 Emails

0.88

URL

0.83

weet

0.82

ronics

0.81

aires

0.81

edin

0.80

Activations Density 0.008%

tweet-related actions or content

references to social media interactions, particularly tweets

No Comments

No Known Activations