INDEX

Explanations

actions related to supporting, helping, or harming others

oai_token-act-pair · gpt-3.5-turbo

actions related to helping, harming, and the ethical implications of those actions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iHUD

-0.77

:]

-0.77

.}

-0.76

 thereof

-0.73

 hers

-0.70

 guiActiveUnfocused

-0.69

................................................................

-0.67

ruff

-0.65

.","

-0.64

ItemThumbnailImage

-0.63

POSITIVE LOGITS

 unsuspecting

1.07

 strangers

0.99

 hordes

0.96

 peoples

0.94

 passers

0.94

 politicians

0.93

 opponents

0.92

 enemies

0.88

 clients

0.88

 celebrities

0.86

Activations Density 0.654%

actions related to supporting, helping, or harming others

actions related to helping, harming, and the ethical implications of those actions

No Comments

No Known Activations