INDEX

Explanations

words related to acts of violence or destruction

oai_token-act-pair · gpt-3.5-turbo

references to violent individuals or those labeled as killers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

uez

-0.75

sit

-0.69

 Instit

-0.68

rely

-0.67

 Marketable

-0.67

oday

-0.66

={

-0.64

nder

-0.63

LECT

-0.63

stall

-0.63

POSITIVE LOGITS

 killer

4.12

 killers

3.26

killer

2.75

 Killer

2.51

 murderer

2.18

killers

1.86

killing

1.81

 murderers

1.79

 assassin

1.68

 rapist

1.59

Activations Density 0.024%

words related to acts of violence or destruction

references to violent individuals or those labeled as killers

No Comments

No Known Activations