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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & 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. 6434
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "converted."

    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
    270
    +0.22
    1.7%
    59
    +0.22
    1.7%
    283
    +0.19
    1.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.22
    0.02
    302
    +0.22
    0.01
    575
    +0.19
    0.02
    Negative Logits
    irtual
    -0.89
    quickShipAvailable
    -0.85
    unta
    -0.84
    otti
    -0.81
    oldown
    -0.78
    utra
    -0.76
    kar
    -0.75
    ongs
    -0.75
    azel
    -0.74
    GV
    -0.74
    POSITIVE LOGITS
    onyms
    0.88
    onym
    0.85
    LEASE
    0.82
     PRESS
    0.80
     Accessed
    0.77
     Faces
    0.76
    xual
    0.76
    ezvous
    0.75
    itbart
    0.73
    ©¶æ
    0.72
    Activations Density 0.025%

    No Known Activations