INDEX

Explanations

mentions of the word "Ale" with varying intensities

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "Ale" in various contexts

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

ä¹ĭ

-0.87

ãĤ¼ãĤ¦ãĤ¹

-0.81

enegger

-0.78

 Everyday

-0.73

ãĥķãĤ©

-0.72

ãģį

-0.68

ãģı

-0.67

circ

-0.66

ãĥķ

-0.65

izations

-0.65

POSITIVE LOGITS

jandro

1.23

Ale

1.07

ppo

0.91

terness

0.89

lette

0.87

rique

0.83

una

0.82

ogan

0.82

Ale

0.81

wine

0.80

Activations Density 0.008%

mentions of the word "Ale" with varying intensities

mentions of the word "Ale" in various contexts

No Comments

No Known Activations