INDEX

Explanations

references to food and cooking concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 27-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_27/width_16k/average_l0_118

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.27.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

GrantedAuthority

-0.51

 vertrouwen

-0.47

 bakom

-0.46

acamata

-0.44

สิน

-0.44

 vété

-0.44

 vrijwilli

-0.43

 podmín

-0.43

 Vertrauen

-0.43

 bezpiecze

-0.42

POSITIVE LOGITS

ReusableCell

0.65

 lunch

0.63

 supper

0.60

 lunchtime

0.60

 dinner

0.57

 rehe

0.53

 lunches

0.53

 sandwich

0.52

 dinners

0.50

lunch

0.50

Activations Density 0.160%

references to food and cooking concepts

No Comments

No Known Activations