INDEX

Explanations

phrases related to clicking or clicking actions

oai_token-act-pair · gpt-3.5-turbo

references to clicking actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Yard

-0.70

 Variety

-0.68

utive

-0.67

 Chancellor

-0.67

 Scotia

-0.64

ZA

-0.64

adium

-0.63

 Minister

-0.61

 Wein

-0.60

Zam

-0.59

POSITIVE LOGITS

lish

0.94

ety

0.90

dress

0.88

through

0.87

wheel

0.87

prints

0.85

pad

0.84

oola

0.75

ername

0.75

click

0.74

Activations Density 0.023%

phrases related to clicking or clicking actions

references to clicking actions

No Comments

No Known Activations