INDEX

Explanations

mathematical concepts related to sequence properties and their classifications

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

approximation properties, interpolating sequences

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

TestingModule

-0.46

CodeAttribute

-0.45

();?>

-0.44

밌

-0.39

randomUUID

-0.39

ScopeManager

-0.39

Yourself

-0.39

WriteLiteral

-0.38

ьаж

-0.38

 насељу

-0.38

POSITIVE LOGITS

 INTERESAR

0.53

 anillos

0.45

دانشنامهٔ

0.43

 συμβ

0.43

 OMITBAD

0.43

 sufficiency

0.42

 resolution

0.42

 intptr

0.42

 concept

0.41

 strategic

0.41

Activations Density 0.633%

mathematical concepts related to sequence properties and their classifications

approximation properties, interpolating sequences

No Comments

No Known Activations