INDEX

Explanations

references to the alphabet

oai_token-act-pair · gpt-3.5-turbo

terms related to the alphabet and its usage

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Spa

-0.76

Tik

-0.72

 Thermal

-0.71

 Buchanan

-0.68

 Releases

-0.67

 Consumption

-0.66

VE

-0.65

MW

-0.65

Ref

-0.65

 Ther

-0.63

POSITIVE LOGITS

 alphabet

4.31

 Alphabet

2.75

phabet

2.12

alph

1.67

 numer

1.31

 chronological

1.22

 acronym

1.22

 acron

1.22

 phon

1.21

 ital

1.20

Activations Density 0.030%

references to the alphabet

terms related to the alphabet and its usage

No Comments

No Known Activations