© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 129688
    Prev
    Next
    INDEX
    Explanations

    task

    np_max-act · gemini-2.0-flash

    The neuron activates on occurrences of the word “task,” i.e. it spotlights that token in instruction prompts.

    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
    row
    -0.06
     nrw
    -0.06
    clo
    -0.06
    ,on
    -0.06
     basil
    -0.06
    mist
    -0.06
    inding
    -0.06
    ,to
    -0.06
    막
    -0.05
    遺
    -0.05
    POSITIVE LOGITS
     stále
    0.07
    fluence
    0.07
    lickr
    0.07
     подход
    0.07
    uestra
    0.07
     dbname
    0.07
    izedName
    0.07
    itere
    0.06
    арод
    0.06
    STATUS
    0.06
    Activations Density 0.003%

    No Known Activations