INDEX

Explanations

words related to embellishments or enhancements

oai_token-act-pair · gpt-3.5-turbo

words that describe the act of making something polished or refined

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 accelerated

-0.68

 hijacked

-0.67

 dictated

-0.66

rogens

-0.66

 supervised

-0.64

agonists

-0.62

 mindful

-0.62

ppard

-0.61

 controlled

-0.60

 reorgan

-0.59

POSITIVE LOGITS

ishes

0.98

ments

0.94

ished

0.85

ãĥĺ

0.83

ings

0.78

enment

0.77

ISH

0.75

âĶĢâĶĢ

0.75

tops

0.75

Interstitial

0.74

Activations Density 0.021%

words related to embellishments or enhancements

words that describe the act of making something polished or refined

No Comments

No Known Activations