INDEX

Explanations

phrases and expressions that indicate uncertainty or conditions surrounding actions and decisions

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

descriptive phrases

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

ніципалі

-0.40

ModelForm

-0.33

+#+

-0.32

AsUp

-0.29

WriteBarrier

-0.28

剩

-0.28

InsertCommand

-0.28

AnimationsModule

-0.28

autoreleasepool

-0.27

itinéraire

-0.26

POSITIVE LOGITS

ftagPool

0.81

ConstraintMaker

0.66

fjspx

0.66

 لينك

0.60

makeConstraints

0.59

 فريبيس

0.58

ɵɵelementStart

0.58

 GenerationType

0.57

loyees

0.57

MLLoader

0.57

Activations Density 0.054%

phrases and expressions that indicate uncertainty or conditions surrounding actions and decisions

descriptive phrases

No Comments

No Known Activations