INDEX

Explanations

terms related to political news and analysis

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

reserved words

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

 propOrder

-0.63

 meisjes

-0.53

AddTagHelper

-0.53

ंदीखरीदारी

-0.52

Personendaten

-0.51

 <<<<<<<<<<<<<<

-0.48

 Hinsicht

-0.47

tangentMode

-0.47

matchCondition

-0.46

IsContent

-0.46

POSITIVE LOGITS

 Formosa

0.47

relief

0.39

anneer

0.38

 виправивши

0.38

Belf

0.38

panel

0.37

楽

0.36

 curtain

0.36

 relief

0.35

sleeping

0.35

Activations Density 0.174%

terms related to political news and analysis

reserved words

No Comments

No Known Activations