INDEX

Explanations

references to notable quotes and their authors

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

attributed quotes

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

 kasarigan

-0.73

GIVEREF

-0.70

Tikang

-0.70

 يتيمه

-0.69

 Chwiliwch

-0.68

 endblock

-0.68

 ब्रेकडाउन

-0.68

 rhestr

-0.67

TagMode

-0.66

ſicht

-0.65

POSITIVE LOGITS

 quotes

0.40

quoted

0.38

 quoted

0.38

Quotes

0.34

Quote

0.33

 quote

0.33

WEET

0.32

baw

0.30

Ho

0.30

miş

0.30

Activations Density 0.015%

references to notable quotes and their authors

attributed quotes

No Comments

No Known Activations