INDEX

Explanations

technical terms and references related to programming and web development

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

technical and descriptive terms

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

⤹

-0.39

 Boulogne

-0.36

CWE

-0.36

+#+#

-0.35

appé

-0.35

rungsseite

-0.34

RegistryLite

-0.34

 Monfieur

-0.34

ceğine

-0.33

 препратки

-0.33

POSITIVE LOGITS

にほんブログ村

0.54

FromNib

0.52

исленность

0.50

AccessorTable

0.50

 crossorigin

0.46

DockStyle

0.45

 gyhoeddwyd

0.44

ISupport

0.42

offsetof

0.41

 bucal

0.41

Activations Density 0.242%

technical terms and references related to programming and web development

technical and descriptive terms

No Comments

No Known Activations