INDEX

Explanations

names of political figures and government bodies

oai_token-act-pair · gpt-3.5-turbo

the end of a document or the document delimiter

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 joking

-0.74

 ______

-0.72

 brill

-0.71

EStream

-0.71

 pandemonium

-0.67

 Rebels

-0.64

 humorous

-0.64

 telling

-0.64

lihood

-0.63

rhy

-0.63

POSITIVE LOGITS

DS

1.28

WS

1.23

SA

1.23

DC

1.22

DP

1.20

ISA

1.20

HS

1.18

LS

1.18

CDC

1.17

1.14

Activations Density 0.208%

names of political figures and government bodies

the end of a document or the document delimiter

No Comments

No Known Activations