INDEX

Explanations

proper nouns, specifically the name "Nate"

oai_token-act-pair · gpt-3.5-turbo

references to the name "Nate."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

omething

-0.72

-0.69

OWER

-0.68

othal

-0.67

hips

-0.67

externalActionCode

-0.67

Ú

-0.67

 Citiz

-0.65

ODUCT

-0.65

inventoryQuantity

-0.65

POSITIVE LOGITS

 Nate

1.06

 Silver

0.89

caster

0.82

laus

0.81

elson

0.81

 Cohn

0.80

elsen

0.79

 Diaz

0.77

 Schmidt

0.75

 Lawson

0.74

Activations Density 0.006%

proper nouns, specifically the name "Nate"

references to the name "Nate."

No Comments

No Known Activations