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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 13523
    Prev
    Next
    INDEX
    Explanations

    references to battles and conflicts, particularly in historical or fictional 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
    383
    +0.19
    1.3%
    385
    +0.18
    1.3%
    268
    +0.17
    1.2%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    268
    +0.19
    0.01
    717
    +0.18
    0.02
    197
    +0.17
    0.02
    Negative Logits
    inki
    -0.80
    ulty
    -0.76
     dstg
    -0.72
    ython
    -0.68
    tan
    -0.67
    krit
    -0.67
    ydia
    -0.65
     slit
    -0.64
    izabeth
    -0.64
    phis
    -0.64
    POSITIVE LOGITS
     Roses
    0.77
    MENTS
    0.74
     predec
    0.73
    MENT
    0.71
    frey
    0.71
     fathers
    0.68
    ¥µ
    0.68
    Players
    0.65
    enaries
    0.65
     Awakens
    0.64
    Activations Density 0.199%

    No Known Activations