INDEX

Explanations

references to political candidates and elections

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

candidate 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

]")]

-0.46

 kurtka

-0.40

baomidou

-0.39

 Photocase

-0.39

間に

-0.39

 monasterio

-0.38

ativement

-0.37

hoea

-0.36

̣c

-0.35

typeDefinition

-0.35

POSITIVE LOGITS

 Candidate

0.62

Candidate

0.61

 ModelExpression

0.60

candidate

0.59

 Prospective

0.57

 candidate

0.57

 Candidates

0.56

 CANDIDATE

0.55

 calon

0.53

didSet

0.52

Activations Density 0.024%

references to political candidates and elections

candidate for office

No Comments

No Known Activations