INDEX

Explanations

phrases referring to specific periods or situations in time

oai_token-act-pair · gpt-3.5-turbo

references to "times" indicating 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 @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ramid

-0.71

 Flavoring

-0.71

Ô

-0.70

RD

-0.68

estead

-0.66

VE

-0.64

aceous

-0.64

Reviewer

-0.63

andan

-0.63

atto

-0.62

POSITIVE LOGITS

cale

1.64

cape

1.15

hare

1.11

pring

1.09

hower

1.06

eries

1.05

pan

1.03

mith

0.94

heet

0.91

creen

0.89

Activations Density 0.032%

phrases referring to specific periods or situations in time

references to "times" indicating various contexts or situations

No Comments

No Known Activations