INDEX

Explanations

the word "Til"

oai_token-act-pair · gpt-3.5-turbo

technical terms and measurements related to experimental physics

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

SQL

-0.79

roup

-0.72

haw

-0.70

é¾į

-0.66

Browser

-0.66

cer

-0.65

 Allaah

-0.62

Maker

-0.62

reader

-0.62

DragonMagazine

-0.62

POSITIVE LOGITS

Til

2.34

 neut

2.18

til

1.37

1.24

 Treat

1.03

til

0.89

 skew

0.86

 Lange

0.84

 Jessie

0.81

 stray

0.77

Activations Density 0.030%

the word "Til"

technical terms and measurements related to experimental physics

No Comments

No Known Activations