INDEX

Explanations

the word "blink" and similar variations

oai_token-act-pair · gpt-3.5-turbo

terms related to visual cues and events, particularly "blink" and its variations

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

amon

-0.79

 Seym

-0.73

 dens

-0.73

AJ

-0.72

amer

-0.71

cum

-0.71

iant

-0.69

isance

-0.69

 bour

-0.68

fou

-0.68

POSITIVE LOGITS

 Blink

1.87

 blink

1.51

 Ripple

1.42

 Matrix

1.41

 Hole

1.37

 Sense

1.30

 Lotus

1.30

 Guns

1.24

 Gravity

1.24

 Gemini

1.23

Activations Density 0.045%

the word "blink" and similar variations

terms related to visual cues and events, particularly "blink" and its variations

No Comments

No Known Activations