INDEX

Explanations

database query structures and parameters

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

name followed by code elements

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

 informée

-0.68

 surla

-0.55

 ExecuteAsync

-0.54

WebVitals

-0.53

 doInBackground

-0.53

StoreMessageInfo

-0.52

InlineData

-0.51

DoubleQuotes

-0.51

UseVisualStyle

-0.50

KommentareTeilen

-0.49

POSITIVE LOGITS

RenderAtEndOf

0.40

SourceChecksum

0.38

 initComponents

0.36

 relevant

0.34

djangoproject

0.34

cik

0.34

Relevant

0.33

tencent

0.32

 basic

0.32

 elog

0.32

Activations Density 0.016%

database query structures and parameters

name followed by code elements

No Comments

No Known Activations