INDEX

Explanations

proper nouns or acronyms related to locations or organizations

oai_token-act-pair · gpt-3.5-turbo

mentions of specific political affiliations or organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

space

-0.72

nell

-0.66

falls

-0.65

blank

-0.65

bar

-0.63

capacity

-0.63

 Fired

-0.62

meyer

-0.62

ted

-0.62

 MacArthur

-0.61

POSITIVE LOGITS

MA

1.02

KING

0.91

ertodd

0.91

uthor

0.89

 opio

0.86

KER

0.81

plin

0.81

BIL

0.80

achusetts

0.79

ignt

0.78

Activations Density 0.006%

proper nouns or acronyms related to locations or organizations

mentions of specific political affiliations or organizations

No Comments

No Known Activations