INDEX

Explanations

terms related to medical procedures and qualifications

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

specific phrases followed by key context words

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

 betweenstory

-0.53

httphttps

-0.47

 يتيمه

-0.45

دانشنامهٔ

-0.44

pageCount

-0.42

 sumpay

-0.41

折

-0.41

野

-0.40

➊

-0.38

 spill

-0.38

POSITIVE LOGITS

addGap

0.50

AnchorStyles

0.47

MemoryWarning

0.46

 Chwiliwch

0.43

ModelSerializer

0.43

Personendaten

0.43

 eléctricas

0.39

 nargin

0.38

 clearfix

0.38

thansa

0.37

Activations Density 0.073%

terms related to medical procedures and qualifications

specific phrases followed by key context words

No Comments

No Known Activations