INDEX

Explanations

terms related to database operations and technical concepts

oai_token-act-pair · gpt-3.5-turbo

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

.ï¿½

-0.59

.''

-0.56

`.

-0.52

lehem

-0.51

cakes

-0.50

ÑĤ

-0.49

''.

-0.49

rade

-0.48

They

-0.48

.</

-0.48

POSITIVE LOGITS

 sanity

0.66

 superhuman

0.65

 sane

0.61

 infall

0.59

 omnip

0.59

 notoriously

0.57

 meticulous

0.56

 ridiculously

0.56

 subconscious

0.56

 infinitely

0.56

Activations Density 1.524%

terms related to database operations and technical concepts

No Comments

No Known Activations