INDEX

Explanations

phrases or words related to origins or original versions

oai_token-act-pair · gpt-3.5-turbo

references to the term "origin."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

âķĲâķĲ

-0.74

lain

-0.69

 Rothschild

-0.65

\\\\\\\\

-0.65

#$

-0.64

 Ventura

-0.64

 fixation

-0.63

vez

-0.63

Case

-0.62

CLASSIFIED

-0.61

POSITIVE LOGITS

inals

1.54

inating

1.53

inations

1.43

inated

1.42

ination

1.40

inally

1.40

inate

1.31

inates

1.29

inators

1.19

inator

1.18

Activations Density 0.056%

phrases or words related to origins or original versions

references to the term "origin."

No Comments

No Known Activations