INDEX

Explanations

keywords related to programming concepts and structures

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

Player Mark

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

package

-0.32

 Wirk

-0.30

astéroïdes

-0.29

備

-0.29

いつか

-0.28

 kabul

-0.28

wart

-0.28

oblastoma

-0.27

Puedes

-0.27

#!

-0.27

POSITIVE LOGITS

Personendaten

0.87

 kasarigan

0.84

Autoritní

0.81

PreInfinity

0.78

WebElementEntity

0.78

 ModelExpression

0.77

GOTREF

0.70

OCCURRED

0.70

UserScript

0.70

 otomatig

0.69

Activations Density 0.054%

keywords related to programming concepts and structures

Player Mark

No Comments

No Known Activations