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. 130551
    Prev
    Next
    INDEX
    Explanations

    diving

    np_max-act · gemini-2.0-flash

    The neuron is detecting occurrences of “dive” and its inflected forms (e.g. diving, dive).

    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
    (Post
    -0.08
     warmth
    -0.07
     McK
    -0.06
     Campbell
    -0.06
    ��态
    -0.06
     کش
    -0.06
    agnar
    -0.06
     stream
    -0.06
     MAIL
    -0.06
     WK
    -0.06
    POSITIVE LOGITS
     diving
    0.13
     dive
    0.13
     Dive
    0.10
     dives
    0.09
     divers
    0.08
     Dove
    0.07
    _div
    0.07
    -div
    0.07
    вы
    0.07
     divisor
    0.07
    Activations Density 0.003%

    No Known Activations