INDEX

Explanations

phrases that indicate distance or extent of concerning matters

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

as far as

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

gonic

-0.50

 ancak

-0.34

 éduc

-0.33

 pluie

-0.33

 nôtre

-0.32

 gritos

-0.32

 jaula

-0.32

 descuentos

-0.31

 bañera

-0.31

 comedor

-0.30

POSITIVE LOGITS

 Infórmanos

0.97

setVerticalGroup

0.90

oweit

0.85

 autorytatywna

0.80

 AssemblyTitle

0.78

 հղումներ

0.77

 linkovi

0.75

addGap

0.74

 CreateTagHelper

0.71

NameInMap

0.69

Activations Density 0.008%

phrases that indicate distance or extent of concerning matters

as far as

No Comments

No Known Activations