INDEX

Explanations

programming and game-related keywords and syntax

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

Integer parsing, game levels

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

IntoConstraints

-0.58

queryInterface

-0.57

 hObject

-0.52

 ब्रेकडाउन

-0.51

évaluateur

-0.50

 Numerade

-0.49

 Signalez

-0.48

IUrlHelper

-0.47

Personendaten

-0.46

 désolés

-0.46

POSITIVE LOGITS

 Levels

0.42

 reduce

0.41

 levels

0.41

Level

0.39

ducir

0.38

Levels

0.36

 kill

0.36

 gameplay

0.36

 lives

0.35

 shrink

0.35

Activations Density 0.003%

programming and game-related keywords and syntax

Integer parsing, game levels

No Comments

No Known Activations