INDEX

Explanations

words related to specific names and surnames

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals or prominent names

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

 Carbuncle

-0.76

 hypot

-0.70

fare

-0.69

nesses

-0.68

cies

-0.68

ebook

-0.67

 Izan

-0.66

xual

-0.63

guards

-0.63

 Raider

-0.63

POSITIVE LOGITS

elsen

1.47

 Nielsen

1.04

Numbers

0.87

wald

0.78

elson

0.75

acher

0.75

Vision

0.74

Limit

0.72

obook

0.72

Ni

0.72

Activations Density 0.008%

words related to specific names and surnames

references to specific individuals or prominent names

No Comments

No Known Activations