INDEX

Explanations

words related to specific individuals, like names

oai_token-act-pair · gpt-3.5-turbo

the letter 'k' 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 @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 mosqu

-0.91

 emergencies

-0.72

 behavi

-0.68

 constitu

-0.67

ãĥ¯

-0.67

 contraceptives

-0.67

 liberties

-0.66

wcsstore

-0.66

 Malf

-0.65

 decay

-0.63

POSITIVE LOGITS

ansas

1.29

irk

1.15

idding

1.15

rieg

1.13

orea

1.10

won

1.01

itty

1.00

icker

0.98

ota

0.95

bps

0.92

Activations Density 0.038%

words related to specific individuals, like names

the letter 'k' in various contexts

No Comments

No Known Activations