INDEX

Explanations

terms and phrases related to cooking and baking instructions

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

Petri dish with

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

 haremos

-0.40

투

-0.38

 Throwable

-0.36

 dessutom

-0.36

 quoque

-0.35

 lãng

-0.34

 advantage

-0.34

astanza

-0.34

 συμβ

-0.34

 attiv

-0.33

POSITIVE LOGITS

 للاسماء

0.53

 محفوظة

0.52

featureID

0.51

rungsseite

0.51

 gynhyrchwyd

0.50

 EdgeInsets

0.49

HasAnnotation

0.48

 الرياضيه

0.46

IUrlHelper

0.46

IBOutlet

0.45

Activations Density 0.016%

terms and phrases related to cooking and baking instructions

Petri dish with

No Comments

No Known Activations