INDEX

Explanations

references to items or objects that are worn out or need replacement

oai_token-act-pair · gpt-3.5-turbo

variations of the words "wear" and "worn."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

aminer

-0.83

emort

-0.69

":["

-0.69

rieved

-0.68

 COUR

-0.65

à¼

-0.65

Ĥ¬

-0.65

yrinth

-0.64

gdala

-0.63

akespeare

-0.63

POSITIVE LOGITS

 worn

1.09

 apparel

0.92

ables

0.86

 wearable

0.84

 clothing

0.83

 Wear

0.82

wear

0.81

bones

0.80

IGH

0.80

 wear

0.79

Activations Density 0.019%

references to items or objects that are worn out or need replacement

variations of the words "wear" and "worn."

No Comments

No Known Activations