INDEX

Explanations

phrases describing something positively as 'nice'

oai_token-act-pair · gpt-3.5-turbo

expressions of positivity and satisfaction

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Estimates

-0.71

-0.70

âĢ

-0.70

uary

-0.68

 recalls

-0.65

omb

-0.62

 miscar

-0.61

 Examination

-0.61

 Chronic

-0.61

borne

-0.61

POSITIVE LOGITS

 nice

3.25

nice

2.39

 nicer

2.23

Nice

2.03

 lovely

2.02

 neat

1.81

 nifty

1.81

 Nice

1.79

 nicely

1.65

 classy

1.60

Activations Density 0.022%

phrases describing something positively as 'nice'

expressions of positivity and satisfaction

No Comments

No Known Activations