Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 130675
    Prev
    Next
    INDEX
    Explanations

    Common English words

    np_max-act · gemini-2.0-flash

    This neuron selectively activates on the characteristic “explanatory” wording used when giving a sentiment‐analysis justification—phrases like “The use of “X” suggests…” and other meta‐commentary around quoted words.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_11/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.11.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    วง
    -0.07
     philosophy
    -0.07
     assistance
    -0.07
     occasion
    -0.07
    rin
    -0.06
     Born
    -0.06
    Communication
    -0.06
     formatter
    -0.06
    born
    -0.06
    理
    -0.06
    POSITIVE LOGITS
     mevcut
    0.07
    /Grid
    0.06
     इसल
    0.06
    ghi
    0.06
     피
    0.06
    _mC
    0.06
     jde
    0.06
    <nav
    0.06
     \`
    0.06
     مان
    0.06
    Activations Density 0.043%

    No Known Activations