INDEX

Explanations

proper nouns or specific entities mentioned in a text

oai_token-act-pair · gpt-3.5-turbo

instances of the word "name" indicating references or mentions of names or entities

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

ysc

-0.79

yrinth

-0.78

isoft

-0.76

qua

-0.74

EMS

-0.69

 Antar

-0.69

*=-

-0.68

Cu

-0.66

zees

-0.66

 Mond

-0.65

POSITIVE LOGITS

 names

0.93

plates

0.91

names

0.84

 naming

0.81

essee

0.76

 tags

0.76

lessly

0.72

 aliases

0.72

calling

0.70

 recognition

0.68

Activations Density 0.025%

proper nouns or specific entities mentioned in a text

instances of the word "name" indicating references or mentions of names or entities

No Comments

No Known Activations