INDEX

Explanations

numbers and numerical formats, particularly in relation to coding or data structures

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

xhtml, xaml, atom, orcid

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_32/width_131k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.32.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

 carreira

-0.29

 taşı

-0.29

 wacht

-0.28

榄

-0.27

 步骤

-0.27

 Barri

-0.27

lauf

-0.26

 Kamin

-0.26

 bilo

-0.25

 dormir

-0.25

POSITIVE LOGITS

цездатний

0.79

IUrlHelper

0.69

 CreateTagHelper

0.65

tagHelperRunner

0.62

 Roskov

0.60

ValueStyle

0.60



0.60

IsMutable

0.60

HideFlags

0.60

findpost

0.59

Activations Density 0.007%

numbers and numerical formats, particularly in relation to coding or data structures

xhtml, xaml, atom, orcid

No Comments

No Known Activations