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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 44777
    Prev
    Next
    INDEX
    Explanations

    in

    np_max-act · gemini-2.0-flash

    The neuron selectively activates on tokens naming labeled steps, shells, or other numbered elements (e.g. “step 2b,” “K-shell,” “K electron”) in technical figure or process descriptions.

    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
    Folders
    -0.06
     tyto
    -0.06
    .order
    -0.06
     ioutil
    -0.06
    mapper
    -0.06
     Messenger
    -0.06
    Both
    -0.06
     desirable
    -0.06
     RS
    -0.06
    /devices
    -0.06
    POSITIVE LOGITS
     confident
    0.07
     Austrian
    0.07
    avs
    0.06
     lieutenant
    0.06
     Anthrop
    0.06
     Απ
    0.06
     �
    0.06
     reconstruct
    0.06
     Tribal
    0.06
    standard
    0.06
    Activations Density 0.036%

    No Known Activations