INDEX

Explanations

mathematical expressions and symbols related to functions, parameters, and characteristics in formal notations

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

Preceding "Tr," "W," "d", "mod", "rk", or "co"

np_acts-logits-general · gemini-2.0-flash

$\mathrm{Tr}$ operations in equations

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

</blockquote>

-0.71

 ever

-0.70

OLOGIES

-0.66

 Rosenthal

-0.64

est

-0.61

ist

-0.59

SaveVideo

-0.59

):=\

-0.59

Coordenadas

-0.59

<?

-0.59

POSITIVE LOGITS

 AppColors

0.87

 Negroes

0.85

Ork

0.82

HX

0.82

 Deity

0.81

eqn

0.80

 wrists

0.79

 Swedes

0.79

 Italij

0.79

EEU

0.79

Activations Density 0.265%

mathematical expressions and symbols related to functions, parameters, and characteristics in formal notations

Preceding "Tr," "W," "d", "mod", "rk", or "co"

$\mathrm{Tr}$ operations in equations

No Comments

No Known Activations