INDEX

Explanations

terms related to hoarding behavior or actions

oai_token-act-pair · gpt-3.5-turbo

terminology related to safeguarding and hoarding

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Plug

-0.68

 confession

-0.62

ÃĹ

-0.60

 confessions

-0.58

 Pain

-0.58

Vo

-0.57

 license

-0.57

Act

-0.57

cro

-0.57

Sun

-0.57

POSITIVE LOGITS

arding

4.48

arded

4.11

ards

2.91

ardless

2.41

ARDS

2.24

ard

2.17

ARD

1.99

ardo

1.45

arde

1.40

arden

1.23

Activations Density 0.013%

terms related to hoarding behavior or actions

terminology related to safeguarding and hoarding

No Comments

No Known Activations