INDEX

Explanations

light-related words

oai_token-act-pair · gpt-3.5-turbo

references to light and illumination

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

ettings

-0.91

isner

-0.73

ESE

-0.69

remlin

-0.67

ĳå£«

-0.67

ocide

-0.65

CHA

-0.64

hov

-0.63

EMENT

-0.63

ese

-0.62

POSITIVE LOGITS

 bulb

1.13

light

1.07

lights

1.04

 bulbs

1.03

weights

0.97

 rays

0.96

lighting

0.94

enment

0.92

 light

0.92

hearted

0.87

Activations Density 0.017%

**light-related** words

references to light and illumination

No Comments

No Known Activations

light-related words