INDEX

Explanations

scientific terminologies or suffixes related to substances or compounds

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

ending in "es"

np_acts-logits-general · gemini-2.0-flash

names ending in es or ces

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

PerformLayout

-0.98

 Monfieur

-0.98

 autorytatywna

-0.97

 purpoſe

-0.92

Personensuche

-0.89

 समीक्षक

-0.89

AndEndTag

-0.89

 незавершена

-0.86

CloseOperation

-0.85

 Signalez

-0.83

POSITIVE LOGITS

es

0.83

ES

0.65

mes

0.63

nes

0.58

sie

0.57

sies

0.56

Kes

0.56

nes

0.55

 TRES

0.54

MES

0.53

Activations Density 0.162%

scientific terminologies or suffixes related to substances or compounds

ending in "es"

names ending in es or ces

No Comments

No Known Activations