© 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. 89710
    Prev
    Next
    INDEX
    Explanations

    taxes and income

    np_max-act · gemini-2.0-flash

    The neuron selectively activates on tokens containing digits—e.g. list indices (“1.”, “2.”) and numeric identifiers (form numbers like “1099”, “W-2”)—i.e. it detects numeric expressions and enumerations.

    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
     offen
    -0.08
    Increasing
    -0.07
    опол
    -0.06
     readOnly
    -0.06
    .Parent
    -0.06
     Ngành
    -0.06
    .@
    -0.06
    apol
    -0.06
    .pa
    -0.06
     اصفه
    -0.06
    POSITIVE LOGITS
    uggested
    0.07
    Conta
    0.06
     lace
    0.06
    _def
    0.06
    JR
    0.06
     warrants
    0.06
     invited
    0.06
     Committee
    0.06
     Sylvia
    0.06
     Mounted
    0.06
    Activations Density 0.004%

    No Known Activations