INDEX

Explanations

references to organizations or groups

oai_token-act-pair · gpt-3.5-turbo

words that indicate possessiveness or belonging

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

 compar

-0.59

 Leilan

-0.57

 Seym

-0.50

¿½

-0.49

SIGN

-0.49

 hemor

-0.49

ļéĨĴ

-0.47

 Lauder

-0.47

EStream

-0.47

EVA

-0.47

POSITIVE LOGITS

pecially

0.75

ELF

0.73

ources

0.69

outhern

0.67

lightly

0.67

omew

0.67

outheast

0.66

aying

0.66

issy

0.64

atisf

0.63

Activations Density 0.213%

references to organizations or groups

words that indicate possessiveness or belonging

No Comments

No Known Activations