INDEX

Explanations

personal pronouns followed by verbs indicating an action or state

oai_token-act-pair · gpt-3.5-turbo

references to subjects performing actions in a narrative

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

math

-0.70

 Cliff

-0.68

 Hayden

-0.65

 Medline

-0.64

icio

-0.64

 Vale

-0.61

 Anat

-0.61

idge

-0.60

 Gerard

-0.60

Geo

-0.60

POSITIVE LOGITS

 encount

0.88

©¶æ

0.83

 frequ

0.81

Ĥİ

0.80

 encountered

0.80

've

0.78

 deems

0.74

ÃĥÃĤ

0.72

 redes

0.71

 owned

0.71

Activations Density 0.192%

personal pronouns followed by verbs indicating an action or state

references to subjects performing actions in a narrative

No Comments

No Known Activations