INDEX

Explanations

terms related to programming and debugging concepts

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

memory usage

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

 preſent

-0.52

 ThemeData

-0.52

 purpoſe

-0.51

addGap

-0.50

 OFDb

-0.48

 pleaſure

-0.46

 AppColors

-0.45

 NgModule

-0.45

 Conſ

-0.44

 OkHttpClient

-0.43

POSITIVE LOGITS

AddHtmlAttribute

0.55

 <=",

0.43

 besk

0.40

 AssemblyCompany

0.38

liyor

0.36

ponses

0.36

 nawet

0.36

⁦

0.35

 debole

0.35

 varandra

0.35

Activations Density 0.106%

terms related to programming and debugging concepts

memory usage

No Comments

No Known Activations