INDEX

Explanations

references to numbers and quantities

oai_token-act-pair · gpt-3.5-turbo

phrases pertaining to measurements, evaluations, or assessments

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

ogether

-0.59

anwhile

-0.53

[*

-0.51

âķĲ

-0.49

 versa

-0.47

 UNCLASSIFIED

-0.45

 moreover

-0.45

âĶ

-0.44

)."

-0.44

 aforementioned

-0.43

POSITIVE LOGITS

 Belfast

0.55

 frontman

0.49

 Transgender

0.49

 cryptocurrency

0.49

 Raspberry

0.49

 Donald

0.49

 graphene

0.48

 transgender

0.46

Bitcoin

0.46

 Grateful

0.46

Activations Density 2.159%

references to numbers and quantities

phrases pertaining to measurements, evaluations, or assessments

No Comments

No Known Activations