INDEX

Explanations

phrases that begin with "In" followed by a year

oai_token-act-pair · gpt-3.5-turbo

references to the structure of written works

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

 endors

-0.70

 lodged

-0.68

 behav

-0.67

 nailed

-0.66

 outwe

-0.65

 Rohing

-0.64

 combatants

-0.64

 unemploy

-0.64

 distur

-0.62

 revoked

-0.62

POSITIVE LOGITS

ventory

1.45

cluding

1.41

strument

1.33

cluded

1.26

clude

1.26

visible

1.26

vasive

1.22

juries

1.22

sect

1.22

iti

1.22

Activations Density 0.134%

phrases that begin with "In" followed by a year

references to the structure of written works

No Comments

No Known Activations