© 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
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 12542
    Prev
    Next
    INDEX
    Explanations

    pre-existing conditions

    np_max-act · gemini-2.0-flash

    The neuron strongly responds to occurrences of the root “exist” and its morphological variants (e.g. “existent,” “existing,” “pre‐existing,” etc.).

    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
    卢
    -0.07
    paragus
    -0.07
     часто
    -0.06
     TELE
    -0.06
     giveaway
    -0.06
     Tele
    -0.06
    ACH
    -0.06
    iele
    -0.06
    ach
    -0.06
     ﾘ
    -0.06
    POSITIVE LOGITS
    validators
    0.07
     gł
    0.07
    -state
    0.07
    Pane
    0.06
     robbery
    0.06
     excessive
    0.06
     وی
    0.06
     необходим
    0.06
    人は
    0.06
    .setData
    0.06
    Activations Density 0.176%

    No Known Activations