© 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
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 109320
    Prev
    Next
    INDEX
    Explanations

    comes/come

    np_max-act · gemini-2.0-flash

    This neuron fires on the discourse pivot “comes” (and similar connective cues like “announcement,” “extension,” “action,” “decision,” “proposal,” and transitional “as” or “after”) that signal a shift or introduction of a new point.

    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
    ських
    -0.06
    微笑
    -0.06
    Roman
    -0.06
     changer
    -0.06
    icao
    -0.06
    "',
    -0.06
    uetype
    -0.06
     基
    -0.06
    емон
    -0.06
    wend
    -0.06
    POSITIVE LOGITS
     Podle
    0.07
    manufacturer
    0.07
    _FILE
    0.07
    lep
    0.07
     Url
    0.06
     swiper
    0.06
     hora
    0.06
    益
    0.06
    ála
    0.06
    ?(
    0.06
    Activations Density 0.041%

    No Known Activations