© 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. 38209
    Prev
    Next
    INDEX
    Explanations

    code and data

    np_max-act · gemini-2.0-flash

    This neuron detects instruction phrasing about gathering and including all information obtained through questions (e.g. “information … gained … through questions,” “include all the information”).

    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
    ㅇ
    -0.07
    thead
    -0.07
    .IsAny
    -0.07
     tấn
    -0.07
    .JMenuItem
    -0.06
    ?type
    -0.06
     MediaType
    -0.06
    판
    -0.06
    count
    -0.06
     headings
    -0.06
    POSITIVE LOGITS
    anno
    0.07
     configurable
    0.06
    148
    0.06
    aya
    0.06
    Direct
    0.06
     иму
    0.06
    $_['
    0.06
    ...)↵
    0.06
    icha
    0.06
     러
    0.06
    Activations Density 0.000%

    No Known Activations