INDEX

Explanations

expressions indicating actions or instructions

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

phrases indicating intention or action

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

way to do something

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-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

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

новниш

-0.73

دانشنامهٔ

-0.66

 JpaRepository

-0.64

 ModelExpression

-0.64

 quæ

-0.61

moveToFirst

-0.59

IntoConstraints

-0.58

 pintados

-0.57

EndInit

-0.57

AddAttribute

-0.57

POSITIVE LOGITS

 approach

0.77

Approach

0.68

 logr

0.67

 ensuring

0.67

 Approach

0.65

 getting

0.63

approach

0.62

 gaining

0.61

 Мексичка

0.60

方式

0.59

Activations Density 0.067%

expressions indicating actions or instructions

phrases indicating intention or action

way to do something

No Comments

No Known Activations