INDEX

Explanations

situations where someone is staring intently or with emotion

oai_token-act-pair · gpt-3.5-turbo

instances of the word "stare" or its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

accompan

-0.70

iHUD

-0.69

FER

-0.64

 Paste

-0.62

 Palestin

-0.60

DragonMagazine

-0.59

é¾į

-0.58

GAN

-0.57

 Recomm

-0.57

ocobo

-0.56

POSITIVE LOGITS

dag

1.10

 attent

1.06

 blank

1.04

 silently

0.94

 gaze

0.91

 stare

0.90

 sideways

0.87

 calmly

0.85

 nervously

0.85

 squarely

0.85

Activations Density 0.042%

situations where someone is staring intently or with emotion

instances of the word "stare" or its variations

No Comments

No Known Activations