INDEX

Explanations

words related to locations or organizations in Los Angeles (LA)

oai_token-act-pair · gpt-3.5-turbo

references to the L.A. Times as a source

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

yi

-0.75

enegger

-0.72

=-=-

-0.67

affle

-0.66

Maker

-0.64

ãĥĨãĤ£

-0.64

="/

-0.64

 Tradable

-0.63

 GPIO

-0.63

bh

-0.63

POSITIVE LOGITS

heed

0.79

.,

0.79

 Lumpur

0.78

velop

0.74

UGH

0.73

.?

0.73

scape

0.70

ven

0.67

illard

0.64

 suburb

0.64

Activations Density 0.020%

words related to locations or organizations in Los Angeles (LA)

references to the L.A. Times as a source

No Comments

No Known Activations