INDEX

Explanations

elements related to experiences, emotions, and significant human interactions

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

state of being full

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

idlertid

-0.67

bibfield

-0.66

 للاسماء

-0.64

ﭜ

-0.61

 Waſſer

-0.60

 fashiola

-0.60



-0.60

 müſſen

-0.57

 queſta

-0.57

enablog

-0.56

POSITIVE LOGITS

ImageContext

0.36

だらけ

0.36

filled

0.33

IBOutlet

0.33

 filled

0.32

full

0.32

 kaynağından

0.32

 contradictions

0.30

ful

0.29

 full

0.29

Activations Density 0.023%

elements related to experiences, emotions, and significant human interactions

state of being full

No Comments

No Known Activations