© 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. 89786
    Prev
    Next
    INDEX
    Explanations

    requests or instructions

    np_max-act · gemini-2.0-flash

    This neuron activates on words relating to instructions or directives—e.g. “follow,” “instructions,” “given,” “provided.”

    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
     forget
    -0.06
     keyPressed
    -0.06
    Bridge
    -0.06
    sequelize
    -0.06
     beat
    -0.06
    anou
    -0.06
     sense
    -0.06
     ['.
    -0.06
     dex
    -0.06
    Animation
    -0.06
    POSITIVE LOGITS
     Bundesliga
    0.07
    _Min
    0.07
    _MT
    0.07
    oci
    0.06
    ]:=
    0.06
    lemn
    0.06
     Austral
    0.06
    ,cp
    0.06
    clearfix
    0.06
     ổn
    0.06
    Activations Density 0.111%

    No Known Activations