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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 127371
    Prev
    Next
    INDEX
    Explanations

    punctuation

    np_max-act · gemini-2.0-flash

    The neuron is keyed to detect list-item prefixes—specifically the numbered or bullet-style markers (e.g. “1.”, “2.”, etc.) that denote entries in a list.

    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
    اکی
    -0.07
    Sem
    -0.06
     reject
    -0.06
     참
    -0.06
    سم
    -0.06
     không
    -0.06
    _BOTTOM
    -0.06
     Spirit
    -0.06
    ATURE
    -0.06
    ине
    -0.06
    POSITIVE LOGITS
    наслід
    0.07
    MeshPro
    0.07
    .nn
    0.07
     ngOn
    0.07
    infer
    0.06
    Оп
    0.06
    }".
    0.06
     малень
    0.06
    .currentThread
    0.06
    學
    0.06
    Activations Density 0.053%

    No Known Activations