© 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. 3491
    Prev
    Next
    INDEX
    Explanations

    patterns related to conditional actions or requirements

    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
    509
    +0.10
    0.4%
    713
    +0.10
    0.4%
    540
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    41
    +0.10
    0.06
    246
    +0.10
    0.06
    521
    +0.10
    0.05
    Negative Logits
    iphate
    -0.88
    ever
    -0.81
    IRT
    -0.76
    athe
    -0.75
    ibly
    -0.75
    uckland
    -0.73
    arter
    -0.72
    egal
    -0.72
    efer
    -0.70
    augh
    -0.69
    POSITIVE LOGITS
     Sync
    0.80
     Mayhem
    0.73
     Compass
    0.71
     Carnival
    0.71
     Expedition
    0.70
     Tempest
    0.70
     Contains
    0.69
     Compliance
    0.68
     Notice
    0.67
     Tribunal
    0.66
    Activations Density 0.294%

    No Known Activations