INDEX

Explanations

programming-related terms and constructs

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

code programming examples

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

findpost

-0.57

EndContext

-0.50

IBOutlet

-0.46

 szóci

-0.46

stånd

-0.46

toprule

-0.46

BeginContext

-0.45

 NgModule

-0.44

 imprimée

-0.43

strcpy

-0.42

POSITIVE LOGITS

 Италијани

0.46

hermes

0.43

 daß

0.42

PopupMenu

0.42

gam

0.42

 approach

0.40

Rema

0.39

eta

0.39

approach

0.37

 approche

0.37

Activations Density 0.018%

programming-related terms and constructs

code programming examples

No Comments

No Known Activations