INDEX

Explanations

references to architectural design and features

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

place and arrangement

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

principalTable

-0.34

 prisión

-0.33

 bénéfices

-0.32

berfläche

-0.32

forschung

-0.32

 frontera

-0.30

 Rptr

-0.29

 navire

-0.29

倒

-0.29

 hogar

-0.29

POSITIVE LOGITS

 CreateTagHelper

0.57

 queſta

0.54

stylers

0.50

 AssemblyTitle

0.50

AddAttribute

0.49

 <>",

0.48

تفصیلات

0.48

saraba

0.47

ftagPool

0.47

conti

0.47

Activations Density 0.018%

references to architectural design and features

place and arrangement

No Comments

No Known Activations