INDEX

Explanations

specific programming and mathematical constructs, particularly related to functions and variables

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

longest measuring

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

BufferException

-0.54

Personendaten

-0.50

挽

-0.45

 auffi

-0.44

 Anſ

-0.41

Slut

-0.41

어

-0.40

Обра

-0.39

icksburg

-0.39

€/

-0.38

POSITIVE LOGITS

 AssemblyCulture

0.60

jsonwebtoken

0.57

enterOuterAlt

0.52

 Normdatei

0.49

IBOutlet

0.48

 groei

0.47

transQ

0.46

Autoritní

0.45

 informée

0.43

цездатний

0.42

Activations Density 0.001%

specific programming and mathematical constructs, particularly related to functions and variables

longest measuring

No Comments

No Known Activations