INDEX

Explanations

return statements in code snippets

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

conditional returns

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

 Henson

-0.28

例文帳に追加

-0.28

 gants

-0.27

 Torneo

-0.26

 meleg

-0.25

 defensor

-0.25

 langage

-0.25

 Mensch

-0.25

iconCls

-0.25

crizione

-0.24

POSITIVE LOGITS

存于互联网档案馆

0.71

UnusedPrivate

0.61

rbrakk

0.61

 disambiguazione

0.60

<unused28>

0.60

<unused51>

0.59

<unused52>

0.59

<unused41>

0.59

<unused3>

0.59

<unused17>

0.59

Activations Density 0.010%

return statements in code snippets

conditional returns

No Comments

No Known Activations