INDEX

Explanations

patterns related to mathematical expressions or programming constructs

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

independent symbol

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

 Normdatei

-0.52

tagHelperRunner

-0.50

 pinulongan

-0.48

WebElementEntity

-0.45

 Exactos

-0.44

 Paglinawan

-0.42

Билгалдахарш

-0.42

queryInterface

-0.42

UrlResolution

-0.41



-0.40

POSITIVE LOGITS

ArrowToggle

0.51

EqualsAnd

0.44

 Beauchamp

0.43

phorbia

0.43

lijker

0.43

oretical

0.43

 Hanley

0.41

esteld

0.41

 George

0.41

Lau

0.40

Activations Density 0.308%

patterns related to mathematical expressions or programming constructs

independent symbol

No Comments

No Known Activations