INDEX

Explanations

quantifiable variables or data points typically used in numerical representations or measurements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 23-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_23/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.23.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

 поши

-0.34

tison

-0.33

 rozší

-0.33

 Konsequ

-0.32

ScopeManager

-0.32

Rptr

-0.32

 läm

-0.32

ifier

-0.32

sled

-0.31

 augmentation

-0.31

POSITIVE LOGITS

AndEndTag

0.75

 Autorizaciones

0.59

 argint

0.58

SharedCtor

0.58

 مشين

0.57

RefNanny

0.57

IntoConstraints

0.56

setVerticalGroup

0.56

MigrationBuilder

0.55

detectChanges

0.55

Activations Density 0.927%

quantifiable variables or data points typically used in numerical representations or measurements

No Comments

No Known Activations