INDEX

Explanations

mentions of the word "Kilner"

oai_token-act-pair · gpt-3.5-turbo

references to the name "Kil."

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

 cheat

-0.67

 breathe

-0.66

 prompt

-0.65

 breathing

-0.62

 prov

-0.62

rant

-0.61

 popup

-0.61

ive

-0.59

 breathed

-0.59

uff

-0.59

POSITIVE LOGITS

Kil

4.06

kil

1.65

kil

1.53

 Rath

1.43

 Fitz

1.23

Kel

1.19

Kin

1.19

Kul

1.17

Kol

1.10

Vil

1.09

Activations Density 0.016%

mentions of the word "Kilner"

references to the name "Kil."

No Comments

No Known Activations