INDEX

Explanations

scientific and technical terms related to parameters and conditions in various contexts

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

mathematical notation with indices

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

 auroit

-0.47

 ſtate

-0.42

 feroit

-0.42

 ainfi

-0.42

 étoit

-0.40

votes

-0.40

 fhew

-0.40

 purpoſe

-0.39

 faſt

-0.39

 Púb

-0.39

POSITIVE LOGITS

 Савезне

0.71

Gambas

0.59

 мәкал

0.57

 autorytatywna

0.56

AndEndTag

0.56

Välislingid

0.54

 PeEnEo

0.53

LabelTagHelper

0.52

alone

0.49

الحياه

0.49

Activations Density 0.230%

scientific and technical terms related to parameters and conditions in various contexts

mathematical notation with indices

No Comments

No Known Activations