INDEX

Explanations

medical or healthcare-related terminology and concepts, particularly references to therapies and interventions

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

Acronyms and abbreviations in parentheses

np_acts-logits-general · gemini-2.0-flash

acronyms ending with specific codes

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

-0.38

 Hör

-0.36

お

-0.35

new

-0.35

Far

-0.35

 Take

-0.34

 take

-0.34

やし

-0.33

ոյ

-0.33

POSITIVE LOGITS

 يتيمه

1.21

 betweenstory

1.16

脚注の使い方

1.13

 '\\;'

1.12

Personensuche

1.11

 дописавши

1.10

 snippetHide

1.08

 فريبيس

1.05

 صوتيه

1.04

 שוליים

1.02

Activations Density 0.229%

medical or healthcare-related terminology and concepts, particularly references to therapies and interventions

Acronyms and abbreviations in parentheses

acronyms ending with specific codes

No Comments

No Known Activations