INDEX

Explanations

numerical values prefixed with a hyphen, indicating negative correlation or decrement

oai_token-act-pair · gpt-3.5-turbo

references to durations of time measured in weeks or longer

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Lanka

-0.71

 cous

-0.62

 Nero

-0.62

Ins

-0.61

 Borders

-0.61

 Rica

-0.60

 commentary

-0.60

 Richards

-0.60

 Claus

-0.59

 Proc

-0.59

POSITIVE LOGITS

year

1.48

month

1.38

digit

1.34

legged

1.32

fif

1.31

minute

1.28

dimensional

1.27

bedroom

1.25

quarters

1.24

week

1.24

Activations Density 0.054%

numerical values prefixed with a hyphen, indicating negative correlation or decrement

references to durations of time measured in weeks or longer

No Comments

No Known Activations