INDEX

Explanations

base keywords associated with specific roles and categories in a structured format

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

sentence ending punctuation

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

참고

-0.64

Cordialement

-0.51

########.

-0.50

jdk

-0.43

chau

-0.43

Sklici

-0.43

ResumeLayout

-0.43

 universitaria

-0.42

cordova

-0.42

inely

-0.41

POSITIVE LOGITS

 Мексичка

0.64

 ویکی‌پدی

0.50

Geografie

0.44

setVerticalGroup

0.43

ThroughAttribute

0.42

WriteTagHelper

0.42

TemporalType

0.42

 noDo

0.41

AddTagHelper

0.40

GIVEREF

0.40

Activations Density 0.150%

base keywords associated with specific roles and categories in a structured format

sentence ending punctuation

No Comments

No Known Activations