INDEX

Explanations

words related to a central location or position

oai_token-act-pair · gpt-3.5-turbo

references to "center" and "centre" as well as related terms indicating a focal point

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 Qiao

-0.65

 Ging

-0.59

Wad

-0.59

ABE

-0.58

Bye

-0.57

IGHTS

-0.57

 Shiny

-0.56

ulously

-0.55

çīĪ

-0.55

 Wool

-0.55

POSITIVE LOGITS

pieces

1.50

piece

1.21

 fielder

1.16

field

1.14

fold

1.08

line

0.99

point

0.99

fire

0.98

stage

0.96

fields

0.91

Activations Density 0.032%

words related to a central location or position

references to "center" and "centre" as well as related terms indicating a focal point

No Comments

No Known Activations