INDEX

Explanations

programming related terms or tools

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the intended use or application of technologies and tools

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

daq

-0.85

é¾

-0.84

atican

-0.81

edin

-0.81

beat

-0.80

aan

-0.77

rose

-0.77

alan

-0.74

soDeliveryDate

-0.74

ouf

-0.74

POSITIVE LOGITS

 detecting

1.47

 storing

1.38

 extracting

1.30

 determining

1.29

 accessing

1.27

 managing

1.24

 converting

1.22

 generating

1.20

 assessing

1.20

 locating

1.19

Activations Density 0.129%

programming related terms or tools

phrases indicating the intended use or application of technologies and tools

No Comments

No Known Activations