INDEX

Explanations

mentions of 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 @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

CLUD

-0.66

 Gent

-0.64

 Myth

-0.64

vantage

-0.63

 Brill

-0.62

 Starr

-0.61

inav

-0.60

 Somers

-0.60

 Valkyrie

-0.60

 Newport

-0.59

POSITIVE LOGITS

storms

1.14

Tweet

1.02

storm

0.99

 Tweet

0.99

 tweets

0.93

 hashtag

0.90

 tweet

0.89

weet

0.87

deck

0.87

 hasht

0.85

Activations Density 0.018%

mentions of tweets

occurrences of the word "tweet"

No Comments

No Known Activations