INDEX

Explanations

phrases containing the word "and."

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

and followed by a name

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

 mijne

-0.37

 taşı

-0.37

 aktiviteter

-0.36

 kív

-0.35

 comportements

-0.35

 avanzado

-0.34

zelfde

-0.34

 that

-0.34

 semejante

-0.34

 mogelijkheid

-0.33

POSITIVE LOGITS

 Administrativna

0.66

تقاوى

0.64

RenderAtEndOf

0.61

 مرئيه

0.61

 disponibilités

0.60

 <>",

0.59

InjectAttribute

0.57

RTDA

0.56

PyExc

0.56

Życiorys

0.56

Activations Density 0.004%

phrases containing the word "and."

and followed by a name

No Comments

No Known Activations