INDEX

Explanations

commands or instructions related to calculations and evaluations

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

evaluate ((

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

 يتيمه

-0.57

MigrationBuilder

-0.56

ImageContext

-0.47

DebuggerStep

-0.45

 termica

-0.44

IContainer

-0.43

enumi

-0.43

flashdata

-0.42

存于互联网档案馆

-0.41

 стоковая

-0.41

POSITIVE LOGITS

 onCreateView

0.45

 Mostly

0.40

 mostly

0.40

########.

0.39

czegó

0.37

 dizer

0.37

mostly

0.37

倌

0.37

Mostly

0.37

どれも

0.37

Activations Density 0.000%

commands or instructions related to calculations and evaluations

evaluate ((

No Comments

No Known Activations