INDEX

Explanations

proper nouns related to people's names

oai_token-act-pair · gpt-3.5-turbo

words specifically related to the term "ole" or variations thereof

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

izoph

-0.74

ultan

-0.73

ulously

-0.72

lished

-0.70

é¾

-0.68

ensical

-0.67

erent

-0.65

wcsstore

-0.64

è¦ļéĨĴ

-0.63

arb

-0.63

POSITIVE LOGITS

tto

1.37

tta

1.28

tti

1.28

cule

1.17

lette

1.12

cules

1.08

cular

0.97

ttes

0.87

tes

0.81

mn

0.81

Activations Density 0.021%

proper nouns related to people's names

words specifically related to the term "ole" or variations thereof

No Comments

No Known Activations