INDEX

Explanations

references to the word "Islam."

oai_token-act-pair · gpt-3.5-turbo

the repeated use of the name "Lamb" or similar references to the same subject throughout the document

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Subtle

-0.69

 Occupations

-0.69

Wan

-0.69

 Slave

-0.64

ãģ¦

-0.63

 Morales

-0.63

 Flavoring

-0.62

Oak

-0.60

ISS

-0.59

LEASE

-0.59

POSITIVE LOGITS

borgh

1.20

ingo

1.15

inate

1.05

ming

1.05

inated

1.04

orously

1.03

beer

1.02

ydia

0.99

inating

0.98

inators

0.98

Activations Density 0.016%

references to the word "Islam."

the repeated use of the name "Lamb" or similar references to the same subject throughout the document

No Comments

No Known Activations