© 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. 3496
    Prev
    Next
    INDEX
    Explanations

    inquiries and reflections about design choices and roles

    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
    634
    +0.13
    0.6%
    368
    +0.12
    0.5%
    493
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    634
    +0.13
    0.14
    446
    +0.12
    0.11
    602
    +0.10
    0.11
    Negative Logits
    Interstitial
    -1.01
    interstitial
    -0.88
     Thumbnails
    -0.84
    thumbnails
    -0.82
    20439
    -0.81
    uploads
    -0.72
     Cosponsors
    -0.72
    irds
    -0.71
    Tips
    -0.70
    itsch
    -0.69
    POSITIVE LOGITS
    ?,
    0.81
     magistrate
    0.79
    Answer
    0.73
     accuser
    0.71
     fate
    0.71
     probable
    0.70
     wiser
    0.70
     heir
    0.68
     attacker
    0.68
    ?),
    0.67
    Activations Density 11.261%

    No Known Activations