INDEX

Explanations

references to products and their features

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

game call holder

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

 μὲν

-0.29

Linq

-0.28

 habet

-0.28

 quidem

-0.26

 Jenderal

-0.25

 eigentlichen

-0.25

 Penelitian

-0.25

qtype

-0.23

 Pembangunan

-0.23

 actual

-0.23

POSITIVE LOGITS

RegistryLite

0.75

tvguidetime

0.74

AndEndTag

0.73

 AssemblyCulture

0.70

 aDecoder

0.69

 intptr

0.68

majánló

0.67

OGND

0.67

saraba

0.66

CppMethod

0.66

Activations Density 0.023%

references to products and their features

game call holder

No Comments

No Known Activations