INDEX

Explanations

mentions of the name "Phil"

oai_token-act-pair · gpt-3.5-turbo

references to the name "Phil" 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 @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

tenance

-0.93

CLASSIFIED

-0.82

yrinth

-0.78

DragonMagazine

-0.71

æĢ

-0.68

Demand

-0.67

CONCLUS

-0.65

shi

-0.65

Rated

-0.63

 Highlander

-0.63

POSITIVE LOGITS

anthrop

1.32

istine

1.16

aimon

1.14

onite

1.08

oshenko

1.06

iton

1.03

adel

1.03

oton

0.99

 Phil

0.99

agraph

0.98

Activations Density 0.006%

mentions of the name "Phil"

references to the name "Phil" in various contexts

No Comments

No Known Activations