INDEX

Explanations

phrases related to specific units of time

oai_token-act-pair · gpt-3.5-turbo

instances of the word "Four" or other numerical references indicating time or quantity

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

 guiActiveUnfocused

-0.75

Ïī

-0.74

oper

-0.72

 surroundings

-0.70

employment

-0.68

à¼

-0.68

ifice

-0.67

 conve

-0.64

ordering

-0.64

omo

-0.64

POSITIVE LOGITS

teen

1.44

een

1.18

eteen

0.97

 Ways

0.95

 Months

0.95

Thirty

0.90

resa

0.88

 Weeks

0.86

fif

0.84

 Hours

0.84

Activations Density 0.103%

phrases related to specific units of time

instances of the word "Four" or other numerical references indicating time or quantity

No Comments

No Known Activations