INDEX

Explanations

phrases related to being constrained or obligated

oai_token-act-pair · gpt-3.5-turbo

phrases that emphasize constraints or limitations

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Soda

-0.75

achu

-0.72

acco

-0.71

Technology

-0.68

ayn

-0.67

 Siem

-0.66

wagen

-0.66

flash

-0.65

emer

-0.64

Website

-0.64

POSITIVE LOGITS

 bound

4.01

bound

2.17

Bound

2.07

 Bound

1.87

 bind

1.76

 binding

1.67

 binds

1.51

 constrained

1.44

 bounds

1.42

binding

1.38

Activations Density 0.008%

phrases related to being constrained or obligated

phrases that emphasize constraints or limitations

No Comments

No Known Activations