© 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. 69167
    Prev
    Next
    INDEX
    Explanations

    Punctuation

    np_max-act · gemini-2.0-flash

    Same activations, with zeros filtered out: <start> called 0.3203125 arguments 1.2548828125 : 2.744140625 `num_numbers`. 0.27685546875 <end> Explanation of neuron 4 behavior: the main thing this neuron does is find the point in documentation where a function’s parameters are being introduced (e.g. the “takes … arguments:” phrase).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    functions that sort elements based on their frequency of occurrence in a given array.

    oai_token-act-pair · gpt-4o-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
    fabric
    -0.07
     اجرای
    -0.07
    陷
    -0.06
     ],
    -0.06
     readdir
    -0.06
     brochure
    -0.06
     )
↵
↵
    -0.06
     Theo
    -0.06
    >false
    -0.06
    asil
    -0.06
    POSITIVE LOGITS
     nga
    0.06
     neurop
    0.06
     adequ
    0.06
    ์ได
    0.06
    Secret
    0.06
    .INVALID
    0.06
     advertise
    0.06
    annotation
    0.06
     phenomena
    0.06
    processors
    0.06
    Activations Density 0.015%

    No Known Activations