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

    Neuronpedia

    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. 41215
    Prev
    Next
    INDEX
    Explanations

    different perspectives

    np_max-act · gemini-2.0-flash

    The neuron detects contrastive or additive discourse markers—words and short phrases (e.g. “but,” “also,” “a bit of”) that signal shifts or qualifications in sentiment.

    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
    _tag
    -0.07
    ,↵
    -0.06
     کسانی
    -0.06
     knight
    -0.06
     Brand
    -0.06
     river
    -0.06
     agriculture
    -0.06
    軽
    -0.06
     Jeffrey
    -0.06
    toString
    -0.06
    POSITIVE LOGITS
    _PAD
    0.07
    SER
    0.06
    αλύτε
    0.06
     DataType
    0.06
    纳
    0.06
    inan
    0.06
     �
    0.06
     Propel
    0.06
    葡
    0.06
     translates
    0.06
    Activations Density 0.028%

    No Known Activations