INDEX

Explanations

short phrases indicating someone causing harm or damage to others

oai_token-act-pair · gpt-3.5-turbo

conjunctions and phrases that indicate connection or addition

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

etermin

-0.83

iosyncr

-0.67

natureconservancy

-0.66

agame

-0.64

agog

-0.64

peak

-0.64

ocative

-0.64

capacity

-0.63

ibble

-0.63

ophon

-0.62

POSITIVE LOGITS

 proceeded

1.30

 forgot

1.30

 ended

1.19

 didnt

1.16

 didn

1.15

 blew

1.15

 gave

1.12

 stayed

1.12

 then

1.11

 threw

1.09

Activations Density 0.348%

short phrases indicating someone causing harm or damage to others

conjunctions and phrases that indicate connection or addition

No Comments

No Known Activations