INDEX

Explanations

references to the United States (U.S.) in various contexts

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

proper nouns related to countries or political entities

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 TAMADRA

-0.73

 Dickinson

-0.72

ks

-0.69

ãĥ¼ãĥ«

-0.67

ãĥ£

-0.66

ersion

-0.65

umen

-0.65

 HOME

-0.64

 Virt

-0.63

ows

-0.63

POSITIVE LOGITS

taboola

0.93

webkit

0.80

_-

0.78

interstitial

0.75

catentry

0.75

hap

0.75

division

0.75

san

0.72

 charism

0.71

alien

0.68

Activations Density 0.010%

references to the United States (U.S.) in various contexts

proper nouns related to countries or political entities

No Comments

No Known Activations