INDEX

Explanations

mentions of the name "Dion" at varying activations

oai_token-act-pair · gpt-3.5-turbo

references to the name "Dion" and related variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

eele

-0.75

ACTION

-0.73

 mileage

-0.71

hirt

-0.71

letcher

-0.69

OAD

-0.69

à¨

-0.67

ABE

-0.66

nesota

-0.66

ridor

-0.64

POSITIVE LOGITS

ysis

1.22

 Dion

0.95

ys

0.88

aea

0.87

acy

0.84

itate

0.81

yss

0.80

gil

0.78

onna

0.74

ros

0.73

Activations Density 0.012%

mentions of the name "Dion" at varying activations

references to the name "Dion" and related variations

No Comments

No Known Activations