INDEX

Explanations

references to specific names or titles, particularly those starting with "El"

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

"El" followed by a place name

np_acts-logits-general · gemini-2.0-flash

el / El followed by a word

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

convention

-0.52

ANDUM

-0.52

 Nowy

-0.51

 Platforms

-0.51

lacion

-0.51

 Audiodateien

-0.50

streng

-0.50

 platforms

-0.50

 journées

-0.49

 convention

-0.49

POSITIVE LOGITS

 Elisha

0.72

el

0.64

El

0.63

ContentAsync

0.60

BeginContext

0.59

0.58

EL

0.58

IUrlHelper

0.58

iprot

0.57

المشاركات

0.56

Activations Density 0.074%

references to specific names or titles, particularly those starting with "El"

"El" followed by a place name

el / El followed by a word

No Comments

No Known Activations