INDEX

Explanations

mentions of mild terms related to positive outlooks or conditions

oai_token-act-pair · gpt-3.5-turbo

terms related to recovery and rehabilitation

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

icable

-0.63

 Hancock

-0.63

 Squadron

-0.62

MEM

-0.60

 dust

-0.59

 flavour

-0.59

 fragrance

-0.59

port

-0.58

stead

-0.57

 salts

-0.57

POSITIVE LOGITS

utenberg

0.88

arcity

0.86

Ó

0.83

elsen

0.82

ktop

0.80

Interstitial

0.78

assetsadobe

0.74

imens

0.74

umbn

0.72

ickr

0.71

Activations Density 0.000%

mentions of mild terms related to positive outlooks or conditions

terms related to recovery and rehabilitation

No Comments

No Known Activations