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. 51
    Prev
    Next
    INDEX
    Explanations

    il

    np_max-act · gemini-2.0-flash

    The neuron activates on the subtoken “umbil” (as in “umbilical”), i.e. it detects mentions of “umbilical.”

    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
     trưởng
    -0.07
    _quality
    -0.07
     đọ
    -0.06
    خواست
    -0.06
    StackTrace
    -0.06
     Prelude
    -0.06
    گز
    -0.06
     dou
    -0.06
     roadway
    -0.06
    "Do
    -0.06
    POSITIVE LOGITS
     till
    0.06
    敏
    0.06
    	img
    0.06
    im
    0.06
    .asarray
    0.06
    .Where
    0.06
    abil
    0.06
       
    0.06
    Where
    0.06
     opportunities
    0.06
    Activations Density 0.000%

    No Known Activations