INDEX

Explanations

references to people being native English speakers

oai_token-act-pair · gpt-3.5-turbo

references to the term "native" in various contexts

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

issors

-0.76

=-=-=-=-

-0.76

awed

-0.75

urat

-0.74

udence

-0.72

etheus

-0.72

enario

-0.70

awks

-0.69

attr

-0.69

abwe

-0.69

POSITIVE LOGITS

 americ

0.92

born

0.86

 Advertisement

0.85

 Instruments

0.79

spe

0.78

izations

0.75

isation

0.70

ization

0.69

izer

0.69

 tongue

0.68

Activations Density 0.019%

references to people being native English speakers

references to the term "native" in various contexts

No Comments

No Known Activations