INDEX

Explanations

phrases related to recognition or acknowledgment

oai_token-act-pair · gpt-3.5-turbo

prepositions and phrases indicating relationships or associations

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

Cooldown

-0.74

issance

-0.72

roo

-0.67

Redditor

-0.66

inventoryQuantity

-0.62

apple

-0.61

interrupted

-0.61

sic

-0.59

zik

-0.58

laughs

-0.58

POSITIVE LOGITS

 their

1.57

 themselves

1.45

their

1.33

 Their

1.29

 THEIR

1.22

 varying

1.16

 respective

1.14

Their

1.13

 differing

0.94

 various

0.91

Activations Density 0.674%

phrases related to recognition or acknowledgment

prepositions and phrases indicating relationships or associations

No Comments

No Known Activations