INDEX

Explanations

keywords related to categories, languages, and skills in a programming context

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

programming terms and keywords

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

 ब्रेकडाउन

-0.61

delwed

-0.58

GTCX

-0.56

__":

-0.53

__':

-0.50

godic

-0.49

zzleHttp

-0.49

NameInMap

-0.49

 jsPsych

-0.49

 onPostExecute

-0.48

POSITIVE LOGITS

 dieux

0.55

 prêtres

0.44

 角色

0.40

 mères

0.40

 оригіналу

0.39

 inconvénients

0.39

árbol

0.38

 dirigeants

0.38

 informaci

0.37

apimachinery

0.37

Activations Density 0.014%

keywords related to categories, languages, and skills in a programming context

programming terms and keywords

No Comments

No Known Activations