INDEX

Explanations

the word "Jag" with different activations ranging from 8 to 10

oai_token-act-pair · gpt-3.5-turbo

the presence of the name "Jag," indicating a focus on a specific character or entity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

lict

-0.83

uctor

-0.75

xual

-0.75

 comparative

-0.69

 prospective

-0.68

aves

-0.66

 competitive

-0.66

ateral

-0.63

 alternating

-0.63

 tentative

-0.61

POSITIVE LOGITS

Jag

1.32

adish

0.89

isu

0.87

rod

0.84

arna

0.84

vana

0.83

oslav

0.83

 Mata

0.81

ita

0.80

Pak

0.79

Activations Density 0.013%

the word "Jag" with different activations ranging from 8 to 10

the presence of the name "Jag," indicating a focus on a specific character or entity

No Comments

No Known Activations