INDEX

Explanations

years within the text

oai_token-act-pair · gpt-3.5-turbo

references to the year 2017

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

iques

-0.64

hiba

-0.62

ÃŃn

-0.62

avis

-0.61

irez

-0.59

tub

-0.58

inks

-0.57

asms

-0.57

cold

-0.55

xual

-0.55

POSITIVE LOGITS

halla

0.82

 onwards

0.75

âĢ³

0.70

 Winners

0.68

ACA

0.67

WER

0.66

 onward

0.66

bleacher

0.66

0.65

 Remastered

0.65

Activations Density 0.053%

years within the text

references to the year 2017

No Comments

No Known Activations