INDEX

Explanations

governmental terms and entities, such as offices, departments, agencies, and organizations

oai_token-act-pair · gpt-3.5-turbo

references to various government and health organizations

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

venge

-0.65

 upside

-0.65

Interstitial

-0.65

 eternity

-0.64

andem

-0.62

laugh

-0.62

ventus

-0.61

rouse

-0.60

Leary

-0.59

imm

-0.59

POSITIVE LOGITS

 recommends

1.30

 publishes

1.30

 warns

1.28

 advises

1.25

 regulates

1.21

 predicts

1.20

 warned

1.18

 issued

1.14

 considers

1.14

 expects

1.13

Activations Density 0.234%

governmental terms and entities, such as offices, departments, agencies, and organizations

references to various government and health organizations

No Comments

No Known Activations