INDEX

Explanations

mentions of or references to relatives

oai_token-act-pair · gpt-3.5-turbo

references to family or relatives

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

oker

-0.79

 Archdemon

-0.78

ococ

-0.69

forcing

-0.68

Cola

-0.67

 Patton

-0.66

oted

-0.65

awar

-0.64

yx

-0.64

 Dome

-0.63

POSITIVE LOGITS

 relatives

1.03

hips

0.96

hesis

0.93

hetical

0.91

heses

0.86

hetically

0.81

 ancest

0.80

hes

0.78

hood

0.77

ilial

0.77

Activations Density 0.016%

mentions of or references to relatives

references to family or relatives

No Comments

No Known Activations