INDEX

Explanations

expressions of degree or intensity related to evaluations or judgments

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

not quite

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

 foncé

-0.40

 duża

-0.39

 dům

-0.39

IonicModule

-0.39

 Vielzahl

-0.38

 unicornio

-0.38

 duży

-0.36

 şir

-0.36

 periódico

-0.35

 createSprite

-0.35

POSITIVE LOGITS

ContentAlignment

0.73

djangoproject

0.64

 ujednoznacz

0.63

KommentareTeilen

0.63

RegressionTest

0.57

Zad

0.57

不太

0.57

 CascadeType

0.56

Personendaten

0.55

 settling

0.55

Activations Density 0.006%

expressions of degree or intensity related to evaluations or judgments

not quite

No Comments

No Known Activations