INDEX

Explanations

elements that indicate enjoyment and entertainment in reading experiences

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

entertaining and enjoyable reading

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

ange

-0.33

 fault

-0.33

ニュ

-0.32

 culture

-0.31

onymy

-0.30

alnız

-0.28

 simb

-0.28

 needing

-0.28

 Equation

-0.28

POSITIVE LOGITS

 entertaining

0.73

 enjoyable

0.69

 entertain

0.68

 Vergnügen

0.66

 entertained

0.66

 Unterhaltung

0.63

RegressionTest

0.63

entertainment

0.62

 divertimento

0.62

 diverti

0.62

Activations Density 0.018%

elements that indicate enjoyment and entertainment in reading experiences

entertaining and enjoyable reading

No Comments

No Known Activations