INDEX

Explanations

references to the word "Kn" followed by a number, potentially related to a specific term or concept

oai_token-act-pair · gpt-3.5-turbo

the term "Knock" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

quo

-0.74

geist

-0.67

 Uran

-0.64

 Alonso

-0.62

 Catalonia

-0.62

Gan

-0.62

ional

-0.61

 Gong

-0.59

REDACTED

-0.59

Gum

-0.59

POSITIVE LOGITS

uckle

1.49

uckles

1.31

ocking

1.28

ighth

1.21

icker

1.19

obs

1.17

itting

1.17

ickers

1.16

ivable

1.15

ows

1.15

Activations Density 0.025%

references to the word "Kn" followed by a number, potentially related to a specific term or concept

the term "Knock" and its variations

No Comments

No Known Activations