INDEX

Explanations

words related to substances or products

oai_token-act-pair · gpt-3.5-turbo

terms related to substances or materials, particularly in a negative or cautionary context

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

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

ItemImage

-0.68

groups

-0.68

physical

-0.67

rons

-0.66

aido

-0.66

 Nass

-0.65

character

-0.65

pdf

-0.64

profile

-0.64

ruct

-0.63

POSITIVE LOGITS

 salesman

1.15

 cans

1.12

 bottle

1.10

 bottles

1.07

 soaked

1.01

 addict

1.01

 vending

0.99

 commercials

0.98

bowl

0.98

 stains

0.97

Activations Density 0.228%

words related to substances or products

terms related to substances or materials, particularly in a negative or cautionary context

No Comments

No Known Activations