INDEX

Explanations

the letter 'V'

oai_token-act-pair · gpt-3.5-turbo

the letter "V" 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

--------------------------------------------------------

-0.83

ãĥ³ãĤ¸

-0.78

ãĤ´ãĥ³

-0.76

éĹĺ

-0.74

¥µ

-0.71

bender

-0.69

ģĸ

-0.68

¿½

-0.67

manship

-0.67

Ú

-0.66

POSITIVE LOGITS

III

1.30

ancouver

1.23

arsity

1.15

alkyrie

1.14

ideos

1.11

ampire

1.11

ascular

1.11

irus

1.10

apor

1.09

iolet

1.09

Activations Density 0.021%

the letter 'V'

the letter "V" in various contexts

No Comments

No Known Activations