INDEX

Explanations

markers or terms related to categorization or identification

oai_token-act-pair · gpt-3.5-turbo

terms related to markers or indicators 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 @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ONSORED

-0.83

é¾

-0.80

æĸ¹

-0.78

erest

-0.78

orld

-0.74

ILLE

-0.74

ŃĶ

-0.73

ategory

-0.72

irl

-0.72

ÄŁ

-0.71

POSITIVE LOGITS

 marker

1.30

 markers

1.25

posts

1.01

 marking

0.83

 plaque

0.73

 indicating

0.73

 pens

0.72

 Carney

0.71

 dotted

0.70

 Twain

0.70

Activations Density 0.010%

markers or terms related to categorization or identification

terms related to markers or indicators in various contexts

No Comments

No Known Activations