INDEX

Explanations

proper nouns related to a specific location, particularly the "Oval Office."

oai_token-act-pair · gpt-3.5-turbo

references to the Oval Office

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

TOR

-0.80

HCR

-0.73

unks

-0.71

ictions

-0.70

 delinquent

-0.68

ARP

-0.67

CVE

-0.67

icia

-0.67

 GOODMAN

-0.66

SOURCE

-0.65

POSITIVE LOGITS

 Oval

0.96

 Office

0.83

lace

0.83

holders

0.73

office

0.72

croft

0.71

 Oath

0.71

bowl

0.70

minster

0.69

cup

0.69

Activations Density 0.009%

proper nouns related to a specific location, particularly the "Oval Office."

references to the Oval Office

No Comments

No Known Activations