INDEX

Explanations

words related to legal entities and names

oai_token-act-pair · gpt-3.5-turbo

references to "Cook" as a significant subject or entity in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 Crusader

-0.70

ilitary

-0.69

ovych

-0.67

emale

-0.65

innie

-0.64

 tremend

-0.64

raints

-0.62

 Pathfinder

-0.62

dfx

-0.62

aundering

-0.62

POSITIVE LOGITS

Cook

1.23

 Cook

1.14

books

0.97

bread

0.94

book

0.94

cook

0.92

batch

0.90

ies

0.89

son

0.88

Gear

0.83

Activations Density 0.007%

words related to legal entities and names

references to "Cook" as a significant subject or entity in various contexts

No Comments

No Known Activations