INDEX

Explanations

contractions of 'do not'

oai_token-act-pair · gpt-3.5-turbo

negations and expressions of inability or refusal

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

 paran

-0.68

anwhile

-0.66

 composite

-0.65

 misdem

-0.65

 princ

-0.64

 retirees

-0.63

 populated

-0.62

 Zoro

-0.61

 ejected

-0.60

ipel

-0.58

POSITIVE LOGITS

Â´

1.33

't

1.28

\'

1.22

ÃŃ

1.19

Â«

1.14

îĢ

1.14

âĶ

1.10

ÃĥÃĤ

1.09

1.08

Activations Density 0.097%

contractions of 'do not'

negations and expressions of inability or refusal

No Comments

No Known Activations