INDEX

Explanations

instances where the word "doesn't" is included

oai_token-act-pair · gpt-3.5-turbo

negations or words indicating absence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

lehem

-0.77

odon

-0.70

igers

-0.65

ende

-0.65

udo

-0.63

 chin

-0.63

zel

-0.63

ghazi

-0.63

iger

-0.63

fall

-0.62

POSITIVE LOGITS

cean

0.81

 charact

0.72

iggurat

0.67

atell

0.65

 corrid

0.65

ospons

0.62

 axis

0.62

hyde

0.62

ajo

0.62

ãĢİ

0.61

Activations Density 0.000%

instances where the word "doesn't" is included

negations or words indicating absence

No Comments

No Known Activations