© 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. 92607
    Prev
    Next
    INDEX
    Explanations

    Grammar/sentences

    np_max-act · gemini-2.0-flash

    The neuron fires strongest on short instructional or section‐heading phrases that cue guidance (e.g. “how to…,” “best way to…,” “Recommend…,” “Connector capabilities”).

    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
     stick
    -0.07
    ẩm
    -0.07
     whose
    -0.06
    REV
    -0.06
    REFIX
    -0.06
     Agriculture
    -0.06
    /Application
    -0.06
     farmer
    -0.06
     Putting
    -0.06
    تها
    -0.06
    POSITIVE LOGITS
    ichern
    0.07
    TableWidgetItem
    0.07
     Değ
    0.06
    -pol
    0.06
     dzi
    0.06
    �
    0.06
     PRIMARY
    0.06
    urb
    0.06
    ukarı
    0.06
    duğ
    0.06
    Activations Density 0.157%

    No Known Activations