INDEX

Explanations

the word "studying"

oai_token-act-pair · gpt-3.5-turbo

terms related to studying or academic pursuits

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

 lapse

-0.81

 vain

-0.70

 arsen

-0.68

anwhile

-0.67

 Algeria

-0.66

 Centauri

-0.66

ocally

-0.66

 capit

-0.65

 anguish

-0.65

 sensing

-0.64

POSITIVE LOGITS

iop

1.08

ios

1.07

ded

1.07

kered

0.95

ents

0.93

sd

0.91

yssey

0.90

ulates

0.89

eenth

0.87

Activations Density 0.025%

the word "studying"

terms related to studying or academic pursuits

No Comments

No Known Activations