INDEX

Explanations

mentions of specific names or individuals in a positive context

oai_token-act-pair · gpt-3.5-turbo

expressions of emotion and connection to others

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 offending

-0.70

 Illegal

-0.68

 incorrectly

-0.67

 delinquent

-0.66

 Restrict

-0.66

error

-0.66

Missing

-0.66

cens

-0.65

 antagonists

-0.65

 censored

-0.63

POSITIVE LOGITS

 grateful

0.96

 gratitude

0.96

 joyful

0.95

 congratulated

0.95

 congratulate

0.94

 ecstatic

0.94

 delighted

0.92

 thrilled

0.91

 prosper

0.91

 compliments

0.91

Activations Density 1.865%

mentions of specific names or individuals in a positive context

expressions of emotion and connection to others

No Comments

No Known Activations