INDEX

Explanations

instances where individuals or groups are being treated in a certain way

oai_token-act-pair · gpt-3.5-turbo

comparisons involving treatment or perception of individuals or groups

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

 Rivals

-0.77

 obser

-0.77

 Riding

-0.74

oln

-0.73

 Archdemon

-0.72

oras

-0.70

chev

-0.69

 Seeking

-0.69

Yon

-0.67

 confir

-0.65

POSITIVE LOGITS

 objectively

0.74

 critically

0.74

 invasive

0.71

 differently

0.71

limits

0.68

ocratic

0.68

bestos

0.67

 atro

0.66

 religiously

0.66

 seriously

0.65

Activations Density 0.094%

instances where individuals or groups are being treated in a certain way

comparisons involving treatment or perception of individuals or groups

No Comments

No Known Activations