Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    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. 40397
    Prev
    Next
    INDEX
    Explanations

    there

    np_max-act · gemini-2.0-flash

    The neuron spikes on the interrogative “Is there any …” phrasing, i.e. the words “there” and “any” when asking if something exists or is possible.

    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
    —are
    -0.08
    —you
    -0.07
    also
    -0.07
    -Up
    -0.07
    _us
    -0.07
    prepared
    -0.06
    _j
    -0.06
    ілі
    -0.06
    -cap
    -0.06
    spin
    -0.06
    POSITIVE LOGITS
     เคร
    0.06
     altın
    0.06
    atican
    0.06
    _countries
    0.06
    Downloader
    0.06
    Frameworks
    0.06
    -Agent
    0.06
     QUESTION
    0.06
     české
    0.06
     اسر
    0.06
    Activations Density 0.011%

    No Known Activations