INDEX

Explanations

words related to specific technological terms or acronyms

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "at" and similar variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

er

-0.66

ADE

-0.65

esses

-0.64

eden

-0.63

 Maxwell

-0.62

ing

-0.60

ed

-0.60

å£«

-0.58

IDER

-0.57

 Tycoon

-0.57

POSITIVE LOGITS

herer

1.45

hered

1.38

chers

1.29

ting

1.22

ters

1.21

tern

1.18

ches

1.17

chell

1.17

chery

1.13

here

1.08

Activations Density 0.090%

words related to specific technological terms or acronyms

occurrences of the word "at" and similar variations

No Comments

No Known Activations