INDEX

Explanations

years, specifically the year 2019

oai_token-act-pair · gpt-3.5-turbo

instances of the year 2019

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

arist

-0.93

riad

-0.87

onym

-0.76

iov

-0.76

onia

-0.74

sonian

-0.73

iqu

-0.70

adata

-0.70

ified

-0.69

arte

-0.68

POSITIVE LOGITS

 onwards

0.99

 onward

0.84

 versa

0.83

0.67

 timeframe

0.66

 inclusive

0.66

 srfAttach

0.65

cible

0.63

å¹

0.62

mAh

0.60

Activations Density 0.005%

years, specifically the year 2019

instances of the year 2019

No Comments

No Known Activations