INDEX

Explanations

phrases and references related to opinions and subjective views

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 33-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_33/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.33.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 purpoſe

-0.55

 Majefty

-0.54

 ſtate

-0.50

ſelves

-0.49

 preſent

-0.48

 Chriftian

-0.48

 poffible

-0.46

 itſelf

-0.46

 ſever

-0.45

 viſ

-0.45

POSITIVE LOGITS

 opinion

3.05

 opinions

2.78

opinion

2.55

 Opinion

2.42

 Opinions

2.36

Opinion

2.30

 opinión

2.27

 opinião

2.20

 OPINION

2.20

 opin

2.17

Activations Density 0.778%

phrases and references related to opinions and subjective views

No Comments

No Known Activations