INDEX

Explanations

actions that can be done without the need for a specific physical item or action

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the necessity or obligation of actions without specific conditions

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

alde

-0.77

DragonMagazine

-0.76

 Vald

-0.74

 Christy

-0.71

vale

-0.67

 Kinnikuman

-0.67

å§«

-0.66

undrum

-0.65

eport

-0.65

roller

-0.64

POSITIVE LOGITS

 whatsoever

0.74

 interruption

0.69

 attribution

0.67

 fuss

0.65

 personally

0.64

 hassle

0.64

 expressly

0.64

 necessarily

0.63

 mention

0.63

 explicitly

0.63

Activations Density 0.143%

actions that can be done without the need for a specific physical item or action

phrases indicating the necessity or obligation of actions without specific conditions

No Comments

No Known Activations