INDEX

Explanations

phrases indicating a position of advantage or progress

oai_token-act-pair · gpt-3.5-turbo

references to being ahead in various contexts or situations

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

Pwr

-0.65

gm

-0.63

Redditor

-0.63

pton

-0.62

ANS

-0.62

uala

-0.60

llular

-0.60

izards

-0.59

rosso

-0.59

tsy

-0.58

POSITIVE LOGITS

of

0.95

paces

0.78

dated

0.77

fitted

0.75

 schedule

0.71

manship

0.71

Of

0.70

 thereof

0.68

wards

0.68

boards

0.67

Activations Density 0.022%

phrases indicating a position of advantage or progress

references to being ahead in various contexts or situations

No Comments

No Known Activations