INDEX

Explanations

phrases related to legal restrictions or prohibitions

oai_token-act-pair · gpt-3.5-turbo

prepositions and phrases indicating exclusion or restriction

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

olitics

-0.88

engine

-0.82

ey

-0.73

arta

-0.73

oiler

-0.71

ISE

-0.71

aic

-0.69

recated

-0.69

alogue

-0.67

inosaur

-0.67

POSITIVE LOGITS

 receiving

1.83

 owning

1.76

 obtaining

1.72

 having

1.70

 accessing

1.69

 registering

1.63

 seeing

1.61

 attending

1.61

 being

1.60

 encountering

1.59

Activations Density 0.606%

phrases related to legal restrictions or prohibitions

prepositions and phrases indicating exclusion or restriction

No Comments

No Known Activations