INDEX

Explanations

programming-related keywords and code structure elements

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

path segments or user context

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

TagMode

-0.81

SharedCtor

-0.69

expandindo

-0.66

afficheront

-0.63

 Taktlose

-0.62

 AssemblyCulture

-0.62

mergeFrom

-0.60

DockStyle

-0.58

MessageOf

-0.55

 noDo

-0.54

POSITIVE LOGITS

 provides

0.43

간

0.42

 consists

0.41

Ent

0.40

 Relationship

0.40

 Sorensen

0.39

 Jansen

0.39

판

0.39

class

0.38

 Faculty

0.38

Activations Density 0.007%

programming-related keywords and code structure elements

path segments or user context

No Comments

No Known Activations