INDEX

Explanations

computer-related terms starting with "DR"

oai_token-act-pair · gpt-3.5-turbo

references to specific product codes or categories in technology

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

 Corpus

-0.90

ropolitan

-0.80

 Pose

-0.77

eers

-0.77

quo

-0.70

 ç¥ŀ

-0.69

geist

-0.66

ria

-0.66

folk

-0.64

 Polic

-0.63

POSITIVE LOGITS

ESS

1.06

adeon

1.00

inks

0.96

INK

0.96

IVE

0.92

OME

0.91

ubs

0.90

IVERS

0.89

inking

0.89

IVER

0.86

Activations Density 0.004%

computer-related terms starting with "DR"

references to specific product codes or categories in technology

No Comments

No Known Activations