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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 91164
    Prev
    Next
    INDEX
    Explanations

    cluster

    np_max-act · gemini-2.0-flash

    This neuron responds to occurrences of the word “Cluster,” especially when it appears as a heading or label.

    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
    804
    -0.07
     ivory
    -0.07
    841
    -0.07
    84
    -0.07
     Ibid
    -0.07
    14
    -0.06
    _eof
    -0.06
    840
    -0.06
    Ford
    -0.06
     Kapoor
    -0.06
    POSITIVE LOGITS
     cluster
    0.17
     Cluster
    0.15
     clusters
    0.14
    cluster
    0.14
    -cluster
    0.11
    Cluster
    0.11
    .cluster
    0.10
    clusters
    0.10
     clustered
    0.09
    (cluster
    0.09
    Activations Density 0.008%

    No Known Activations