INDEX

Explanations

positive descriptors such as 'bright' that convey optimism or positivity

oai_token-act-pair · gpt-3.5-turbo

the word "bright" and its variations, indicating positivity or potential

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

abase

-0.72

 Abuse

-0.71

Wars

-0.70

against

-0.67

olo

-0.66

 Press

-0.66

rm

-0.65

berman

-0.64

 Oral

-0.63

 Attacks

-0.63

POSITIVE LOGITS

 bright

3.50

 brighter

2.69

 brightly

2.20

 brightest

2.12

Bright

1.69

 Bright

1.69

 brightness

1.65

bright

1.59

dim

1.55

 dark

1.50

Activations Density 0.024%

positive descriptors such as 'bright' that convey optimism or positivity

the word "bright" and its variations, indicating positivity or potential

No Comments

No Known Activations