INDEX

Explanations

terms indicating inclusivity or comprehensiveness

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

all items or groups

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

 autorytatywna

-0.61

ReusableCell

-0.61

 Chwiliwch

-0.55

phorbia

-0.53

 Administrativna

-0.52

ServletException

-0.51

 AssemblyProduct

-0.50

 CURLOPT

-0.50

__(/*!

-0.49

 glyphicon

-0.49

POSITIVE LOGITS

ALL

0.72

All

0.67

all

0.62

All

0.59

Others

0.54

 Others

0.52

 trésor

0.50

showAll

0.50

setAll

0.50

 GetAll

0.48

Activations Density 0.011%

terms indicating inclusivity or comprehensiveness

all items or groups

No Comments

No Known Activations