INDEX

Explanations

numerical values and guidelines associated with time or duration

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

Prepositions and words starting with "h"

np_acts-logits-general · gemini-2.0-flash

shall be increased by

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 dodgy

-0.83

Indeed

-0.78

 famously

-0.73

Firstly

-0.71

 pensado

-0.71

 Indeed

-0.70

 Firstly

-0.70

 Sadly

-0.70

Ещё

-0.69

 indeed

-0.68

POSITIVE LOGITS

 hereinafter

0.84

 shall

0.82

 herein

0.76

shall

0.73

 skall

0.72

 Partagez

0.71

 leçons

0.70

 auffi

0.69

Efq

0.69

 ejus

0.69

Activations Density 0.977%

numerical values and guidelines associated with time or duration

Prepositions and words starting with "h"

shall be increased by

No Comments

No Known Activations