INDEX

Explanations

foods or objects that can potentially be harmful when ingested

oai_token-act-pair · gpt-3.5-turbo

various types of food items and edible ingredients

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 practition

-0.86

eatures

-0.82

ONSORED

-0.75

PDATE

-0.75

 lobb

-0.71

 experien

-0.68

CLASS

-0.66

Student

-0.65

 affili

-0.65

ISA

-0.64

POSITIVE LOGITS

poons

1.13

mith

1.09

poon

1.05

creen

1.02

cale

0.98

pin

0.98

pring

0.98

 dipped

0.98

 coated

0.97

ilver

0.94

Activations Density 0.204%

foods or objects that can potentially be harmful when ingested

various types of food items and edible ingredients

No Comments

No Known Activations