INDEX

Explanations

references to suspects in a legal or investigative context

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

suspect

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

Демографія

-0.51

 Maurer

-0.50

 Италијани

-0.49

 Farina

-0.48

 Maire

-0.48

время

-0.47

 Giordano

-0.46

 disambiguazione

-0.46

 maș

-0.46

 Dalio

-0.46

POSITIVE LOGITS

 suspect

0.84

Sus

0.73

Sus

0.69

 suspects

0.66

 Sussex

0.63

 susceptibility

0.62

sus

0.61

SUS

0.60

Susan

0.58

SUS

0.58

Activations Density 0.009%

references to suspects in a legal or investigative context

suspect

No Comments

No Known Activations