INDEX

Explanations

phrases related to moral values and principles

oai_token-act-pair · gpt-3.5-turbo

references to core values and principles

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

geon

-0.76

å¥³

-0.73

--------------------------------------------------------

-0.72

waves

-0.69

fac

-0.68

sie

-0.67

\/\/

-0.64

OVA

-0.63

Sus

-0.63

brain

-0.62

POSITIVE LOGITS

 values

0.82

 ideals

0.79

 Values

0.76

 beliefs

0.75

values

0.75

 propositions

0.74

 proposition

0.74

 principles

0.73

 embodied

0.72

 sensibilities

0.71

Activations Density 0.016%

phrases related to moral values and principles

references to core values and principles

No Comments

No Known Activations