INDEX

Explanations

phrases introducing information or attributing statements to sources

oai_token-act-pair · gpt-3.5-turbo

instances of the word "apparently."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ces

-0.73

 funeral

-0.69

 respir

-0.67

raised

-0.64

alla

-0.64

cum

-0.63

estern

-0.62

ocol

-0.60

andering

-0.59

yrs

-0.59

POSITIVE LOGITS

 Apparently

0.84

Apparently

0.84

Buyable

0.80

NESS

0.70

Magikarp

0.69

ãĥ¼ãĥĨãĤ£

0.68

orious

0.66

SAY

0.66

èĥ

0.65

Known

0.65

Activations Density 0.006%

phrases introducing information or attributing statements to sources

instances of the word "apparently."

No Comments

No Known Activations