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. 84447
    Prev
    Next
    INDEX
    Explanations

    dehyde

    np_max-act · gemini-2.0-flash

    The neuron specializes in detecting the names of chemical reactions and reaction types (e.g. “Mannich reaction,” “aldol condensation,” “Grignard reagent,” etc.).

    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
                        
    ↵
    -0.08
    on
    -0.07
    .pivot
    -0.07
    йом
    -0.07
     intents
    -0.07
     enthusiasm
    -0.06
    坦
    -0.06
    ﺔ
    -0.06
    acker
    -0.06
    _datas
    -0.06
    POSITIVE LOGITS
    dehyde
    0.10
    dehy
    0.08
     Comedy
    0.07
    ном
    0.07
    toMatchSnapshot
    0.07
     blurred
    0.06
    $search
    0.06
     Flem
    0.06
    lerinde
    0.06
    <bits
    0.06
    Activations Density 0.005%

    No Known Activations