INDEX

Explanations

coding-related terms and constructs in programming context

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

in mind,

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

awaiter

-0.69

queryInterface

-0.60

onedDateTime

-0.59

 Roskov

-0.59

 يتيمه

-0.55

évaluateur

-0.54

matchCondition

-0.54

ScopeManager

-0.54

ExtendWith

-0.52

copg

-0.50

POSITIVE LOGITS

 rodillas

0.44

 carcasa

0.42

 prestaciones

0.41

 laiko

0.41

 Geschäfte

0.41

šča

0.40

 macetas

0.40

 arcilla

0.39

 táctil

0.39

ultades

0.39

Activations Density 0.000%

coding-related terms and constructs in programming context

in mind,

No Comments

No Known Activations