INDEX

Explanations

references to political campaigns and candidates

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

running for office

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

ValueStyle

-0.47

ThroughAttribute

-0.45

脚注の使い方

-0.45

expandindo

-0.44

 INVOL

-0.41

 sumpay

-0.41

 <=",

-0.41

 CanadaChoose

-0.41

Diweddarwch

-0.40

.”—

-0.40

POSITIVE LOGITS

 candidate

0.53

 Candidate

0.52

candidate

0.48

Candidate

0.47

 candidacy

0.47

 racer

0.44

 Candidates

0.44

 candidates

0.44

 Campaign

0.43

stateProvider

0.43

Activations Density 0.024%

references to political campaigns and candidates

running for office

No Comments

No Known Activations