INDEX

Explanations

the term "native" in the context of language or culture

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

RFC

-0.73

urat

-0.73

bard

-0.71

pta

-0.70

sbm

-0.70

quer

-0.68

^^^^

-0.67

inventoryQuantity

-0.67

ramid

-0.67

atically

-0.67

POSITIVE LOGITS

 americ

1.10

 Hawai

0.95

 Hawaiian

0.93

 Advertisement

0.90

born

0.86

 tongue

0.86

 speakers

0.82

spe

0.82

 speaker

0.81

 Instruments

0.81

Activations Density 0.033%

the term "native" in the context of language or culture

No Comments

No Known Activations