© 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. 41681
    Prev
    Next
    INDEX
    Explanations

    Technology and robots

    np_max-act · gemini-2.0-flash

    This neuron activates on the sequence “EM” “IEW” plus following digits—that is, it detects the robot model name acronym “EMIEW” with its numeric suffix.

    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
    체
    -0.07
    _work
    -0.07
     bağlantılar
    -0.07
    Awesome
    -0.07
    _dma
    -0.07
     ByteArrayInputStream
    -0.06
    ,rp
    -0.06
    (delay
    -0.06
    Transmission
    -0.06
     cosa
    -0.06
    POSITIVE LOGITS
     turbine
    0.06
    YG
    0.06
     sustain
    0.06
    律宾
    0.06
     recycle
    0.06
    (photo
    0.06
    assing
    0.06
    ibase
    0.06
     resized
    0.06
     '';↵
    0.06
    Activations Density 0.012%

    No Known Activations