INDEX

Explanations

mentions of organizations or groups referred to as "entities"

oai_token-act-pair · gpt-3.5-turbo

references to various types of entities, particularly in legal or organizational contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

stroke

-0.79

geon

-0.71

jin

-0.70

edo

-0.66

éĹĺ

-0.66

afa

-0.65

Progress

-0.64

eful

-0.62

joy

-0.60

cho

-0.60

POSITIVE LOGITS

 entity

1.00

 entities

0.86

chnology

0.83

 owned

0.80

 headquartered

0.75

ional

0.75

 owners

0.68

 controlled

0.68

 owner

0.67

 accelerator

0.67

Activations Density 0.018%

mentions of organizations or groups referred to as "entities"

references to various types of entities, particularly in legal or organizational contexts

No Comments

No Known Activations