INDEX

Explanations

a variety of phrases indicating positivity or agreement

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

so to speak

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

دانشنامهٔ

-0.59

feinander

-0.44

 istore

-0.42

Demografie

-0.41

sendRedirect

-0.40

guém

-0.38

ikirkan

-0.38

 metra

-0.37

جمعیت

-0.37

Across

-0.37

POSITIVE LOGITS

poz

2.59

Poz

1.98

Poz

1.91

poz

1.82

 пози

1.20

 pozy

1.08

пози

0.97

 поз

0.96

 Posi

0.92

ポジ

0.92

Activations Density 0.002%

a variety of phrases indicating positivity or agreement

so to speak

No Comments

No Known Activations