INDEX

Explanations

phrases emphasizing minimal disturbance or unnecessary commotion

oai_token-act-pair · gpt-3.5-turbo

terms related to complaints or disturbances

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ACTED

-0.71

 plane

-0.69

 commute

-0.67

 prison

-0.66

ramer

-0.66

 corridor

-0.66

 Peninsula

-0.64

ramid

-0.64

 prisoner

-0.60

ombs

-0.60

POSITIVE LOGITS

 fuss

1.16

naire

1.08

ãĤ¦ãĤ¹

0.91

iness

0.90

naires

0.90

engers

0.89

 Leilan

0.86

cake

0.86

ĸļ

0.83

eful

0.81

Activations Density 0.008%

phrases emphasizing minimal disturbance or unnecessary commotion

terms related to complaints or disturbances

No Comments

No Known Activations