INDEX

Explanations

code constructs and logical conditions related to boolean checks

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

programming symbols and multilingual communication

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 informée

-0.81

:✨

-0.73

RegressionTest

-0.71

 pinulongan

-0.63

IContainer

-0.61

Билгалдахарш

-0.60

 otomatig

-0.60

 surla

-0.60

__":

-0.59

 المعيارى

-0.59

POSITIVE LOGITS

urlpatterns

0.41

 comunicación

0.30

 piedi

0.30

 Comunicación

0.28

 koning

0.28

उ

0.28

السلام

0.28

 utilización

0.27

 gracios

0.27

CommandType

0.26

Activations Density 0.046%

code constructs and logical conditions related to boolean checks

programming symbols and multilingual communication

No Comments

No Known Activations