INDEX

Explanations

numbers-related terms, specifically numerical values and measurements

oai_token-act-pair · gpt-3.5-turbo

economic terms related to taxation and financial impacts

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

iannopoulos

-0.55

ogie

-0.54

photos

-0.54

inav

-0.53

shaw

-0.52

illas

-0.52

olitics

-0.50

 Rodham

-0.49

lance

-0.48

gif

-0.48

POSITIVE LOGITS

 nevertheless

0.62

 nonetheless

0.61

due

0.56

 administr

0.56

 relying

0.56

 econom

0.55

 excessively

0.55

 blindly

0.54

 unforeseen

0.53

 uncertainties

0.53

Activations Density 2.720%

numbers-related terms, specifically numerical values and measurements

economic terms related to taxation and financial impacts

No Comments

No Known Activations