INDEX

Explanations

references to personal identity and self

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

alignment items

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

تقاوى

-0.63

ReusableCell

-0.58

óbal

-0.49

@[+][

-0.49

balleur

-0.48

 KeyCode

-0.47

 convertView

-0.46

 bogotá

-0.46

 utafitiHapana

-0.46

TypedDataSet

-0.46

POSITIVE LOGITS

alignItems

0.68

items

0.65

 alignItems

0.64

extAlignment

0.59

AxisAlignment

0.46

getMinutes

0.44

complexContent

0.44

ITEMS

0.44

him

0.43

GOTREF

0.43

Activations Density 0.000%

references to personal identity and self

alignment items

No Comments

No Known Activations