INDEX

Explanations

phrases related to quotations or spoken words

oai_token-act-pair · gpt-3.5-turbo

instances of a particular character or symbol in text

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 segreg

-0.75

 Cycl

-0.73

 tremend

-0.72

 assemb

-0.72

 sling

-0.71

 cloak

-0.70

 dispers

-0.69

Alc

-0.67

 scatter

-0.67

 mans

-0.66

POSITIVE LOGITS

ľ

1.64

¦

1.28

¬

1.25

º

1.25

¼

1.24

¤

1.19

Ľ

1.18

Ń

1.18

ª

1.18

¿

1.15

Activations Density 0.231%

phrases related to quotations or spoken words

instances of a particular character or symbol in text

No Comments

No Known Activations