INDEX

Explanations

user interactions and sentiments

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

social media hashtags

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

outSlope

-0.39

 NoSuch

-0.35

titian

-0.32

 treated

-0.32

inigte

-0.32

 насеље

-0.31

udd

-0.31

FormTagHelper

-0.30

 behand

-0.30

Лю

-0.29

POSITIVE LOGITS

 Instagram

0.59

 hashtags

0.58

Instagram

0.56

 tweets

0.54

 instagram

0.54

uxxxx

0.54

 resourceCulture

0.53

tiérrez

0.53

 tweeting

0.53

 hashtag

0.52

Activations Density 0.019%

user interactions and sentiments

social media hashtags

No Comments

No Known Activations