INDEX

Explanations

Twitter account usernames related to various public figures

oai_token-act-pair · gpt-3.5-turbo

names of public figures and organizations

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

 causation

-0.62

blance

-0.62

ordinate

-0.61

 watering

-0.61

 bloated

-0.60

 aggrav

-0.59

 extradition

-0.59

 cumbersome

-0.59

 visitation

-0.57

 faculties

-0.57

POSITIVE LOGITS

(@

1.35

 âĺ

0.70

 âľ

0.69

Hunter

0.69

ische

0.67

Scot

0.67

Parent

0.66

 âĻ

0.65

 Byte

0.64

zeb

0.64

Activations Density 0.113%

Twitter account usernames related to various public figures

names of public figures and organizations

No Comments

No Known Activations