INDEX

Explanations

terminology associated with research and scientific studies

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

future on, public on, developmental steps

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

 للاسماء

-0.73

 للمعارف

-0.69

 myſelf

-0.68

Datuak

-0.66

 kasarigan

-0.66

AddTagHelper

-0.63

withIdentifier

-0.63

 zijne

-0.61

genodigd

-0.61

 queſta

-0.61

POSITIVE LOGITS

cam

0.25

 thats

0.25

 draft

0.24

 desperation

0.24

 Glad

0.24

 among

0.24

 campaign

0.24

 bill

0.24

addGap

0.24

Activations Density 0.319%

terminology associated with research and scientific studies

future on, public on, developmental steps

No Comments

No Known Activations