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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    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. 128953
    Prev
    Next
    INDEX
    Explanations

    pen

    np_max-act · gemini-2.0-flash

    The neuron selectively activates on occurrences of the substring “pen,” whether as the standalone token or as part of words like “penicillin.”

    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
     更
    -0.07
    _HIGH
    -0.07
     segregation
    -0.07
    arda
    -0.07
    arcy
    -0.06
    high
    -0.06
    .Static
    -0.06
    \Validator
    -0.06
    _RA
    -0.06
    .MixedReality
    -0.06
    POSITIVE LOGITS
     Pen
    0.15
     pen
    0.14
    Pen
    0.10
     pens
    0.09
     pencil
    0.09
    pen
    0.09
     penned
    0.09
    _pen
    0.08
     penc
    0.08
    .pen
    0.08
    Activations Density 0.012%

    No Known Activations