INDEX

Explanations

references to arms or related terminology in various contexts

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

arm

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

*/;

-0.42

>`;

-0.40

 Census

-0.40

 zaby

-0.40

RTHOOK

-0.39

detectChanges

-0.39

تج

-0.39

=>

-0.39

Caz

-0.39

</table>

-0.39

POSITIVE LOGITS

Arm

1.17

Arm

1.16

arm

1.09

ARM

1.05

Arms

1.00

arm

0.98

ARM

0.96

 Arms

0.96

 brazo

0.87

 arms

0.86

Activations Density 0.014%

references to arms or related terminology in various contexts

arm

No Comments

No Known Activations