© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Qwen3.5-9B Base
    3. 15-QWENSCOPE-RES-64K
    4. 40700
    Prev
    Next
    INDEX
    Explanations

    entity category has a ManyToOne relation with type so each category has one type..." (mentions category, type) - "... namespace Plato.Ideas.Categories.Models { public class CategoryDetails : Serializable { public int TotalEntities { get; set..." (mentions category, models) - "... public class SysMenuTest implements Serializable{ private String id; private String title; private String href; ..." (mentions menu, test, data structures) - "... Unable to determine the principal end of an association between the types "JobsApp.Job" and "JobsApp.Category"." (mentions category, types) - "... retailscm.trainingcoursetype; ...`trainingcoursetype`..." (mentions type, training) - "... select all categories from a webservice." (mentions category) - "... public class CultureModel { public string Culture { get; set; } ..." (mentions culture model)**Pattern Identification:**The `MAX_ACTIVATING_TOKENS` list prominently features words like `category`, `categories`, `genre`, `Culture`, `Menu`, and `type`.The `TOP_POSITIVE_LOGITS` suggest administrative or structural concepts like `CRUD`, `admin`, `Manage`, `hier` (hierarchy).The `TOP_ACTIVATING_TEXTS` provide context:- "admin_brand_title", "search_list_configuration" - relates to configuration, administration.- "category has a ManyToOne relation with type", "JobsApp.Job" and "JobsApp.Category" - clearly about relationships between entities or data types.- "SysMenuTest" - about menu structures.- "CultureModel" - about culture types.- "trainingcoursetype" - about training types.The common themes are:1. **Data / Entity Categories**: `category`, `genre`, `type`2. **Structural/Navigational**: `Menu`, `Models`3. **Administrative/Configurational**: `admin`, `CRUD`, `configuration`4. **Types/Classifications**: `type`, `category`, `genre`, `Culture`The combination of categories/types with administrative or structural contexts seems to be a strong pattern. The `TOKENS_AFTER_MAX_ACTIVATING_TOKEN` like `_title`, `.Models`, `Model` also reinforce this.The most specific and recurring pattern is the "type" or "category" of things, often in a data model or administrative context.Let's consider some direct explanations:- "entity types and categories" (4 words) - Good, identifies the core elements.- "data types and categories" (4 words) - Similar, good.- "category and type relations" (4 words) - Focuses on relations.- "admin types and categories" (4 words) - Adds admin context.- "menu and category types" (4 words) - Adds menu context.entity types and categories

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    igu
    -0.07
    ietta
    -0.07
    @student
    -0.06
    æ¯ĶåĪ«äºº
    -0.06
    ç¼ł
    -0.06
    otte
    -0.06
    æĹłåıį
    -0.06
    iland
    -0.06
    _signature
    -0.06
    engar
    -0.06
    POSITIVE LOGITS
     CRUD
    0.10
    Crud
    0.08
     hier
    0.08
    管çIJĨ
    0.08
    Hier
    0.08
     admin
    0.08
    çζ
    0.08
    ưỡng
    0.07
    á»īnh
    0.07
    Manage
    0.07
    Activations Density 0.340%

    No Known Activations