© 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. 55795
    Prev
    Next
    INDEX
    Explanations

    interactive processes

    np_max-act · gemini-2.0-flash

    The neuron activates on words used in the meta‐instructions or directives (e.g. “respond,” “inputs,” “message”), so it’s detecting instruction- or direction-related language.

    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
     War
    -0.06
    альним
    -0.06
     Animals
    -0.06
     deutsche
    -0.06
    -0.06
     Bench
    -0.06
    HW
    -0.06
     Races
    -0.06
     але
    -0.06
     EMP
    -0.06
    POSITIVE LOGITS
     footing
    0.07
    κε
    0.07
    .removeItem
    0.07
    	Element
    0.06
    金
    0.06
    Slides
    0.06
    व
    0.06
     residing
    0.06
     getDescription
    0.06
     محمد
    0.06
    Activations Density 0.010%

    No Known Activations