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. 130193
    Prev
    Next
    INDEX
    Explanations

    where

    np_max-act · gemini-2.0-flash

    The main thing this neuron does is detect the discourse‐pivot phrase “this is where,” flagging sentences that introduce an explanatory elaboration.

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

    references to energy management and consulting practices.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    instances where the word "where" is used to introduce or indicate a particular clause or explanation.

    oai_token-act-pair · gpt-4oTriggered 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
     screenplay
    -0.07
     Svg
    -0.07
     Dissertation
    -0.06
    vant
    -0.06
    errmsg
    -0.06
     kinetics
    -0.06
    _equal
    -0.06
    τρο
    -0.06
    _filenames
    -0.06
    ISS
    -0.05
    POSITIVE LOGITS
    こう
    0.07
     edilen
    0.07
     escalated
    0.06
    amız
    0.06
    り
    0.06
     mga
    0.06
    import
    0.06
    Within
    0.06
     vamp
    0.06
    โดย
    0.06
    Activations Density 0.011%

    No Known Activations