INDEX

Explanations

references to engineering competitions and teams

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

design challenge

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

 autorytatywna

-0.41

 nahilalakip

-0.36

ыгана

-0.36

Autoritní

-0.35

Referanser

-0.35

 للاسماء

-0.33

 Wicidata

-0.33

 дописавши

-0.32

 iprot

-0.32

MigrationBuilder

-0.31

POSITIVE LOGITS

 Challenge

0.58

Plug

0.57

 Plug

0.57

 ingenuity

0.56

 challenge

0.55

 Pitch

0.54

 solvers

0.54

Challenge

0.53

isition

0.52

 winners

0.52

Activations Density 0.009%

references to engineering competitions and teams

design challenge

No Comments

No Known Activations