INDEX

Explanations

keywords and terms related to identity and social roles

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

characters or archetypes

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

maxcdn

-0.38

 OkHttpClient

-0.33

porus

-0.32

🌽

-0.32

 externi

-0.31

Handlung

-0.30

 comuniques

-0.30

 igenom

-0.30

 Behaviour

-0.29

件事情

-0.29

POSITIVE LOGITS

SequentialGroup

0.59

GOTREF

0.56

msgTypes

0.56

 Dieſe

0.54

 Nacionales

0.54

 verſch

0.53

<unused14>

0.53

<unused23>

0.53

<unused8>

0.53

<pad>

0.53

Activations Density 0.051%

keywords and terms related to identity and social roles

characters or archetypes

No Comments

No Known Activations