INDEX

Explanations

the letter "M" or words starting with "M" followed by a digit

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the letter 'M' in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Extras

-0.76

ãĤ¡

-0.70

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.69

 Beyond

-0.67

DragonMagazine

-0.66

 yours

-0.65

 Pose

-0.64

 Bleach

-0.63

 Eleven

-0.63

 Samar

-0.62

POSITIVE LOGITS

astern

1.13

ugg

1.13

apping

1.11

oto

1.11

iscal

1.10

igrant

1.08

aim

1.07

akin

1.07

OST

1.07

oses

1.07

Activations Density 0.040%

the letter "M" or words starting with "M" followed by a digit

repeated mentions of the letter 'M' in various contexts

No Comments

No Known Activations