INDEX

Explanations

steps or instructions in a procedural context, particularly related to cooking or technical processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 tartalomajánló

-0.58

 незавершена

-0.58

 cherchés

-0.56

ponses

-0.56

الإنجليزية

-0.56

 pleaſure

-0.56

 Италијани

-0.55

astify

-0.53

WithMany

-0.50

InputBorder

-0.50

POSITIVE LOGITS

Then

0.42

 then

0.42

 Prepare

0.38

ifikationer

0.38

Prepare

0.38

 références

0.38

 cautiously

0.38

 prepare

0.37

 gently

0.37

 Then

0.36

Activations Density 0.307%

steps or instructions in a procedural context, particularly related to cooking or technical processes

No Comments

No Known Activations