INDEX

Explanations

positive information or highlights in text

oai_token-act-pair · gpt-3.5-turbo

phrases that convey positive news or highlights about various topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

ivalent

-0.74

indust

-0.70

adult

-0.70

heit

-0.70

igmatic

-0.68

ancies

-0.67

amental

-0.66

throp

-0.66

urch

-0.64

POSITIVE LOGITS

 bonus

0.71

 avoids

0.67

 cures

0.66

 additions

0.66

 luckily

0.65

:]

0.64

 rewards

0.64

 overlooking

0.63

 cushion

0.63

 Bonus

0.62

Activations Density 0.166%

positive information or highlights in text

phrases that convey positive news or highlights about various topics

No Comments

No Known Activations