INDEX

Explanations

phrases related to expressing personal opinions and preferences

oai_token-act-pair · gpt-3.5-turbo

expressions of personal sentiment and emotional states

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

 omin

-0.66

 commonplace

-0.63

geries

-0.62

 prominence

-0.61

 vacuum

-0.59

 aggregation

-0.57

 flourish

-0.57

 shun

-0.57

 shortages

-0.56

ulton

-0.56

POSITIVE LOGITS

 myself

1.34

âĺ

1.04

 gonna

0.95

 ðŁĳ

0.87

sorry

0.85

ya

0.84

 ðŁĺ

0.84

 thankful

0.84

ãĤ´

0.83

 kidding

0.83

Activations Density 0.361%

phrases related to expressing personal opinions and preferences

expressions of personal sentiment and emotional states

No Comments

No Known Activations