INDEX

Explanations

negative statements or criticisms

oai_token-act-pair · gpt-3.5-turbo

negative constructions and expressions of prohibition

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

 biod

-0.69

soDeliveryDate

-0.68

PD

-0.66

quickShipAvailable

-0.65

LIN

-0.65

Wid

-0.64

ãĤ¼ãĤ¦ãĤ¹

-0.62

 Bonds

-0.62

 veins

-0.61

 Valkyrie

-0.61

POSITIVE LOGITS

 emulate

0.97

erers

0.81

iners

0.72

umbn

0.72

regnancy

0.72

 imitate

0.71

 apologize

0.70

 aspire

0.70

 behave

0.70

icably

0.70

Activations Density 0.057%

negative statements or criticisms

negative constructions and expressions of prohibition

No Comments

No Known Activations