INDEX

Explanations

words related to "meta" concepts or analysis

oai_token-act-pair · gpt-3.5-turbo

references to "meta" concepts or discussions

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

shire

-0.80

Interstitial

-0.74

iens

-0.73

 Franks

-0.72

 Petty

-0.71

 Shooter

-0.70

teenth

-0.68

IBLE

-0.68

ivity

-0.67

 Loyal

-0.65

POSITIVE LOGITS

ethical

1.00

alyses

0.93

plot

0.92

allo

0.85

 guiActiveUn

0.83

hei

0.79

econom

0.77

verse

0.77

enter

0.76

ethyl

0.75

Activations Density 0.016%

words related to "meta" concepts or analysis

references to "meta" concepts or discussions

No Comments

No Known Activations