INDEX

Explanations

person names

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names of individuals and places

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

ModLoader

-0.84

âĶĢâĶĢ

-0.70

etheless

-0.65

duino

-0.64

HMS

-0.62

 Lanka

-0.62

 FANTASY

-0.60

 Imran

-0.59

 Emblem

-0.58

ãĥŁ

-0.57

POSITIVE LOGITS

zinski

0.78

zen

0.78

heny

0.76

lett

0.73

inski

0.73

zik

0.73

linger

0.72

bey

0.72

beck

0.71

zman

0.71

Activations Density 0.254%

person names

proper nouns, particularly names of individuals and places

No Comments

No Known Activations