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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 84693
    Prev
    Next
    INDEX
    Explanations

    semicolon or parenthesis

    np_max-act · gemini-2.0-flash

    This neuron detects code patterns that navigate up a parent/ancestor chain (e.g. uses `.parentNode`, `.parent`, or similar parent‐access calls).

    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
    Disconnect
    -0.07
    holding
    -0.07
    .setBackgroundResource
    -0.07
    resolver
    -0.07
    igkeit
    -0.06
    .setSelection
    -0.06
    싱
    -0.06
     ((*
    -0.06
    SKI
    -0.06
    ама
    -0.06
    POSITIVE LOGITS
     CLL
    0.07
    .pnl
    0.07
     lá
    0.07
     الخام
    0.07
    ')))↵
    0.07
    .News
    0.06
     yarn
    0.06
     pari
    0.06
    "));↵
    0.06
     ])↵↵
    0.06
    Activations Density 0.018%

    No Known Activations