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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsBlogSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Attention Out - 128k
    5. 0-ATT_128K-OAI
    6. 6425
    Prev
    Next
    INDEX
    Explanations

    instances of the word "placed" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    jbloom/GPT2-Small-OAI-v5-128k-attn-out-SAEs/v5_128k_layer_0
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    131,072
    Data Type
    torch.float32
    Hook Name
    blocks.0.hook_attn_out
    Hook Layer
    0
    Architecture
    standard
    Context Size
    64
    Dataset
    Skylion007/openwebtext
    Activation Function
    topk
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    710
    +0.21
    1.5%
    739
    +0.15
    1.1%
    679
    +0.15
    1.1%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    642
    +0.21
    0.00
    276
    +0.15
    0.00
    710
    +0.15
    0.00
    Negative Logits
    externalActionCode
    -0.93
    asis
    -0.71
     Rept
    -0.68
    ebook
    -0.67
    class
    -0.67
    omics
    -0.66
     Warehouse
    -0.64
    MM
    -0.64
    reci
    -0.63
     ---------
    -0.63
    POSITIVE LOGITS
    nesday
    1.07
    rontal
    1.01
    etheless
    0.93
     pse
    0.93
    iculty
    0.92
    icult
    0.87
    ulously
    0.87
     unbeliev
    0.85
    enhagen
    0.82
     srf
    0.80
    Activations Density 0.003%

    No Known Activations