© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Attention Out - 32k
    5. 3-ATT_32K-OAI
    6. 3090
    Prev
    Next
    INDEX
    Explanations

    terms related to organizations and teamwork

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-att_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-attn-out-SAEs/v5_32k_layer_3
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.3.hook_attn_out
    Hook Layer
    3
    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
    13
    +0.11
    0.5%
    464
    +0.11
    0.5%
    7
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    65
    +0.11
    0.01
    464
    +0.11
    0.01
    556
    +0.10
    0.01
    Negative Logits
     ende
    -0.81
     destro
    -0.74
    atio
    -0.73
     tremend
    -0.73
     corrid
    -0.71
    encies
    -0.70
    inse
    -0.70
    GGGGGGGG
    -0.69
    cephal
    -0.69
    nels
    -0.67
    POSITIVE LOGITS
     lost
    0.74
    yrs
    0.68
    erk
    0.66
    '
    0.66
    owan
    0.64
     Gifts
    0.62
     Marketplace
    0.61
     NX
    0.61
     Musical
    0.61
     Confederation
    0.61
    Activations Density 0.054%

    No Known Activations