INDEX

Explanations

phrases related to the upper part of the body

oai_token-act-pair · gpt-3.5-turbo

references to the term "upper" along with related classifications or locations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

-0.78

atile

-0.73

INESS

-0.72

tnc

-0.72

 Machina

-0.70

Deal

-0.68

ulously

-0.68

naire

-0.66

 Chaser

-0.65

ensable

-0.64

POSITIVE LOGITS

most

1.21

 tiers

0.92

class

0.89

 lobe

0.88

lip

0.88

 extrem

0.86

 tier

0.85

 limb

0.85

 layers

0.83

parts

0.81

Activations Density 0.037%

phrases related to the upper part of the body

references to the term "upper" along with related classifications or locations

No Comments

No Known Activations