INDEX

Explanations

keywords related to specific concepts or terms within various contexts, such as legal matters, success/failure judgments, artistic movements, medical conditions, or physical locations

oai_token-act-pair · gpt-3.5-turbo

discussions related to significant societal issues or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Machines

-0.72

doms

-0.67

rencies

-0.67

ernels

-0.65

brids

-0.64

assies

-0.62

lees

-0.61

rogens

-0.60

 channels

-0.60

 Scenes

-0.60

POSITIVE LOGITS

 standpoint

0.79

 contender

0.63

 consisting

0.60

cko

0.57

 approximation

0.56

 perspective

0.56

 apology

0.56

 spokesperson

0.56

 variant

0.55

 staffer

0.55

Activations Density 2.487%

keywords related to specific concepts or terms within various contexts, such as legal matters, success/failure judgments, artistic movements, medical conditions, or physical locations

discussions related to significant societal issues or events

No Comments

No Known Activations