INDEX

Explanations

phrases related to personal opinions or judgments

oai_token-act-pair · gpt-3.5-turbo

inquiries or assertions related to racial issues

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

 respectively

-0.63

EMS

-0.61

 themselves

-0.61

 Trident

-0.55

 Hebdo

-0.51

 apiece

-0.50

 arsen

-0.50

 Hole

-0.50

Their

-0.49

 Autob

-0.49

POSITIVE LOGITS

 myself

1.32

poke

0.82

my

0.70

ividual

0.67

 personally

0.65

cyclopedia

0.62

ograp

0.62

chin

0.60

anmar

0.59

é»Ĵ

0.58

Activations Density 1.099%

phrases related to personal opinions or judgments

inquiries or assertions related to racial issues

No Comments

No Known Activations