© 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. 73705
    Prev
    Next
    INDEX
    Explanations

    together

    np_max-act · gemini-2.0-flash

    The neuron spots occurrences of the word “together” (especially when it begins a coordinating phrase).

    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
    ột
    -0.07
    _ind
    -0.07
    емати
    -0.07
    noho
    -0.07
    MLElement
    -0.07
     leaks
    -0.07
    新闻
    -0.07
    dbuf
    -0.07
    ляють
    -0.06
     않았다
    -0.06
    POSITIVE LOGITS
     Together
    0.10
     together
    0.08
     join
    0.08
     treasure
    0.08
     combined
    0.07
     engaged
    0.07
     taper
    0.06
     spolu
    0.06
     junto
    0.06
     away
    0.06
    Activations Density 0.015%

    No Known Activations