INDEX

Explanations

uncommon characters or symbols

oai_token-act-pair · gpt-3.5-turbo

negative contractions and various uses of the word "it."

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

•GPT2 L5 Resid: RRLLRLLRLLRRRRLLRRRRRRRRR

No Comments

Negative Logits

anwhile

-0.75

 scatter

-0.69

 scattering

-0.68

gad

-0.62

 nonexistent

-0.61

 Saga

-0.61

osate

-0.61

 dangling

-0.60

 nearest

-0.59

 muse

-0.59

POSITIVE LOGITS

º

1.17

£

1.12

¹

1.08

į

0.95

 âĢº

0.94

ı

0.93

§

0.93

®

0.91

¬

0.88

¿

0.88

Activations Density 0.389%

uncommon characters or symbols

negative contractions and various uses of the word "it."

No Comments

No Known Activations