INDEX

Explanations

phrases that summarize or evaluate the subjects positively

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

overall assessment

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

ⓧ

-0.76

LLocation

-0.67

хьтан

-0.67

 يتيمه

-0.66

 ***!

-0.63

الإنجليزية

-0.62

 Italijani

-0.62

 queſta

-0.62

 wireType

-0.62

rungsseite

-0.62

POSITIVE LOGITS

Overall

0.90

 overall

0.90

 Overall

0.87

overall

0.78

总体

0.63

OVERALL

0.61

整體

0.60

整体

0.54

 Insgesamt

0.52

 keseluruhan

0.50

Activations Density 0.006%

phrases that summarize or evaluate the subjects positively

overall assessment

No Comments

No Known Activations