INDEX

Explanations

measurements and instances of quantifying in a document

oai_token-act-pair · gpt-3.5-turbo

instances of the word "measuring" and related concepts

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

cffffcc

-0.82

iferation

-0.75

fml

-0.75

ItemTracker

-0.72

vous

-0.71

 afore

-0.69

 attendant

-0.65

udence

-0.64

she

-0.63

 appearance

-0.63

POSITIVE LOGITS

uate

1.09

olate

0.99

kson

0.80

 yourselves

0.71

resses

0.71

urate

0.70

verning

0.70

ments

0.70

urat

0.68

ribe

0.67

Activations Density 0.043%

measurements and instances of quantifying in a document

instances of the word "measuring" and related concepts

No Comments

No Known Activations