INDEX

Explanations

locations or landmarks followed by the word "National"

oai_token-act-pair · gpt-3.5-turbo

references to national parks and wildlife refuges

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

ullivan

-0.78

qs

-0.74

redd

-0.72

utenberg

-0.72

raq

-0.71

roxy

-0.70

lass

-0.69

enger

-0.68

ylan

-0.67

 cylinders

-0.66

POSITIVE LOGITS

 Geographic

1.05

 Monument

1.03

 Laboratory

1.02

 Laboratories

1.02

 Park

1.00

 Recreation

0.99

 Guard

0.98

 Cemetery

0.96

 Historic

0.95

 Forest

0.93

Activations Density 0.024%

locations or landmarks followed by the word "National"

references to national parks and wildlife refuges

No Comments

No Known Activations