INDEX

Explanations

markers indicating code structure or annotations in programming languages

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

/* code comments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

ValueStyle

-0.60

 محفوظة

-0.57

脚注の使い方

-0.53

 kasarigan

-0.52

ukone

-0.50

 geslacht

-0.48

-0.47

 tartalomajánló

-0.46

Lähteet

-0.46

 AssemblyTitle

-0.45

POSITIVE LOGITS

/*

0.42

 คว

0.39

ayakan

0.38

 gardes

0.37

ParseTree

0.35

ésus

0.35

 handleMessage

0.35

vigil

0.34

 abzu

0.34

 anzub

0.34

Activations Density 0.004%

markers indicating code structure or annotations in programming languages

/* code comments

No Comments

No Known Activations