INDEX

Explanations

references to rankings or lists of top entities

oai_token-act-pair · gpt-3.5-turbo

references to rankings and lists of notable entities or accomplishments

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

displayText

-0.84

bol

-0.78

icum

-0.78

DERR

-0.71

amination

-0.69

uchi

-0.69

amar

-0.67

pty

-0.66

¬¼

-0.66

£ı

-0.64

POSITIVE LOGITS

ottest

1.15

 Worst

1.08

 safest

1.05

 happiest

1.04

 Favorite

1.04

Top

1.01

Top

0.99

 Highest

0.99

 coolest

0.97

 Greatest

0.96

Activations Density 0.260%

references to rankings or lists of top entities

references to rankings and lists of notable entities or accomplishments

No Comments

No Known Activations