INDEX

Explanations

words related to laws, regulations, and legal entities

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 internationally

-0.54

ãĥ¼ãĥĨ

-0.49

 practically

-0.48

 Nickel

-0.48

 beforehand

-0.48

 presumably

-0.48

 Western

-0.47

 Marketplace

-0.47

 UCLA

-0.46

 afterward

-0.46

POSITIVE LOGITS

anus

0.72

ip

0.69

isoft

0.68

Äĵ

0.68

gae

0.67

aser

0.66

ax

0.65

ape

0.65

atin

0.64

asing

0.64

Activations Density 0.587%

words related to laws, regulations, and legal entities

No Comments

No Known Activations