INDEX

Explanations

phrases related to decisions, votes, or actions that are unanimous

oai_token-act-pair · gpt-3.5-turbo

instances of unanimous decisions or agreements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

lla

-0.81

href

-0.77

nels

-0.74

 Extrem

-0.73

resources

-0.72

vas

-0.70

 Gutenberg

-0.70

anza

-0.68

neys

-0.68

gin

-0.67

POSITIVE LOGITS

 unanimously

1.29

 unanimous

1.02

 vote

0.80

 approve

0.79

 voted

0.77

IFIED

0.77

 ratified

0.77

 approving

0.75

 approves

0.75

 minded

0.73

Activations Density 0.012%

phrases related to decisions, votes, or actions that are unanimous

instances of unanimous decisions or agreements

No Comments

No Known Activations