INDEX

Explanations

words related to a video game character named "Duke" in various contexts

oai_token-act-pair · gpt-3.5-turbo

the repeated mention of "Duke Nukem" or related phrases within the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

heny

-0.76

earchers

-0.73

angers

-0.62

 blurred

-0.61

igree

-0.60

 constit

-0.59

 wearer

-0.59

agna

-0.58

raised

-0.57

ifax

-0.56

POSITIVE LOGITS

uke

1.08

UGH

0.90

hyde

0.84

vich

0.80

 Haram

0.75

hiro

0.74

pedia

0.72

ernel

0.72

unin

0.72

oleon

0.71

Activations Density 0.006%

words related to a video game character named "Duke" in various contexts

the repeated mention of "Duke Nukem" or related phrases within the text

No Comments

No Known Activations