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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Attention Out - 32k
    9. 3-ATT_32K-OAI
    11. 4048
    Prev
    Next
    INDEX
    Explanations

    terms related to technical jargon and operational definitions

    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
    430
    +0.11
    0.5%
    452
    +0.10
    0.4%
    445
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.11
    0.05
    586
    +0.10
    0.03
    284
    +0.09
    0.03
    Negative Logits
    oday
    -0.72
    today
    -0.71
     Shroud
    -0.69
     sky
    -0.67
     sunset
    -0.66
     Reloaded
    -0.66
     overhead
    -0.66
    stack
    -0.63
     Restoration
    -0.63
     react
    -0.63
    POSITIVE LOGITS
    hemer
    0.75
    gha
    0.74
    cial
    0.71
    healthy
    0.66
    ciating
    0.64
    ruciating
    0.64
    ionage
    0.64
    vre
    0.62
     Deaths
    0.61
     Kaf
    0.61
    Activations Density 0.451%

    No Known Activations