INDEX

Explanations

words related to Indigenous communities or individuals

oai_token-act-pair · gpt-3.5-turbo

references to Aboriginal and related names or terms

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

osen

-0.78

nut

-0.77

agonist

-0.71

engers

-0.71

akia

-0.71

hof

-0.70

asted

-0.68

role

-0.68

apult

-0.67

ide

-0.67

POSITIVE LOGITS

razil

0.80

herty

0.79

 Cortana

0.75

DX

0.73

 Aboriginal

0.72

 Occupations

0.71

 Saskatchewan

0.71

 McKay

0.69

Footnote

0.68

 Rodney

0.67

Activations Density 0.034%

words related to Indigenous communities or individuals

references to Aboriginal and related names or terms

No Comments

No Known Activations