INDEX

Explanations

phrases indicating ability or capability

oai_token-act-pair · gpt-3.5-turbo

phrases indicating potential or ability

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

cot

-0.75

cott

-0.70

bard

-0.68

Swe

-0.67

meric

-0.67

Moz

-0.67

VEN

-0.65

alter

-0.65

BA

-0.63

mol

-0.63

POSITIVE LOGITS

nesses

0.92

 capable

0.83

 incapable

0.80

bodied

0.78

 impairment

0.77

acers

0.76

icient

0.76

 enough

0.72

sonian

0.72

 selves

0.72

Activations Density 0.011%

phrases indicating ability or capability

phrases indicating potential or ability

No Comments

No Known Activations