INDEX

Explanations

warnings or cautions given about certain actions or situations

oai_token-act-pair · gpt-3.5-turbo

warnings or advisories related to various subjects

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

lez

-0.78

è£

-0.70

ULAR

-0.69

 partName

-0.65

 largeDownload

-0.65

avez

-0.65

oin

-0.64

Laughs

-0.63

alter

-0.62

oyer

-0.62

POSITIVE LOGITS

 lest

0.83

 warnings

0.80

 impending

0.77

irens

0.70

 heights

0.69

 Signs

0.67

wolves

0.67

 dangers

0.65

autions

0.62

warn

0.61

Activations Density 0.128%

warnings or cautions given about certain actions or situations

warnings or advisories related to various subjects

No Comments

No Known Activations