INDEX

Explanations

keywords related to DC Comics

oai_token-act-pair · gpt-3.5-turbo

references to "DC" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

xual

-1.02

lihood

-0.92

htt

-0.84

olulu

-0.78

tery

-0.75

htar

-0.75

ï¸

-0.70

ptin

-0.70

 myster

-0.67

ãĥĭ

-0.67

POSITIVE LOGITS

 Comics

1.05

NF

1.00

DC

0.82

olor

0.80

NM

0.79

ription

0.78

AT

0.77

ascade

0.77

MJ

0.75

ULT

0.73

Activations Density 0.018%

keywords related to DC Comics

references to "DC" in various contexts

No Comments

No Known Activations