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

    Neuronpedia

    Jacobian LensNEW
    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. 56966
    Prev
    Next
    INDEX
    Explanations

    a

    np_max-act · gemini-2.0-flash

    The neuron is picking up on sentence-initial or mid-sentence discourse cues—short adverbs or prepositional phrases (e.g. “Still,” “With a nod,” “With a smirk,” etc.) that transition or frame narrative action.

    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
    ，如
    -0.06
    utter
    -0.06
     Koch
    -0.06
    att
    -0.06
     userInfo
    -0.06
    jc
    -0.06
    <TextView
    -0.06
     metabol
    -0.06
     drawbacks
    -0.06
    _variable
    -0.06
    POSITIVE LOGITS
    ?><
    0.07
    _down
    0.06
    _EXPORT
    0.06
    	me
    0.06
     Covered
    0.06
     Berm
    0.06
    FFT
    0.06
    finger
    0.06
    assed
    0.06
     QVBoxLayout
    0.06
    Activations Density 0.018%

    No Known Activations