INDEX

Explanations

information related to programming languages and localization settings

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

language tags like lang="en"

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

AndEndTag

-0.68

 المعيارى

-0.60

 Biôgrafia

-0.51

WriteTagHelper

-0.49

DockStyle

-0.48

 kaarangay

-0.47

AccessorTable

-0.45

 EconPapers

-0.45

parsedMessage

-0.45

 ſehr

-0.45

POSITIVE LOGITS

 Audiodateien

0.38

MathML

0.38

 Forsch

0.37

 souha

0.37

elt

0.37

atott

0.35

 باب

0.35

にあった

0.34

providedIn

0.34

ilihan

0.34

Activations Density 0.005%

information related to programming languages and localization settings

language tags like lang="en"

No Comments

No Known Activations