INDEX

Explanations

phrases related to exploration, investigation, or medical emergencies

oai_token-act-pair · gpt-3.5-turbo

terms related to exploration and observation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Force

-0.75

 Potion

-0.71

 Shack

-0.71

 Sapphire

-0.68

-0.67

 Dome

-0.65

BALL

-0.65

 Biological

-0.64

 Rebellion

-0.63

 Formation

-0.62

POSITIVE LOGITS

atory

1.53

ations

1.48

ational

1.40

ances

1.37

ants

1.28

ancies

1.22

ables

1.20

estate

1.19

acies

1.13

ant

1.13

Activations Density 0.064%

phrases related to exploration, investigation, or medical emergencies

terms related to exploration and observation

No Comments

No Known Activations