© 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
    3. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 8-RES_POST_32K-OAI
    11. 4305
    Prev
    Next
    INDEX
    Explanations

    specific technical terms and classifications related to electronics and nature

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.8.hook_resid_post
    Hook Layer
    8
    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
    534
    +0.11
    0.5%
    112
    +0.10
    0.5%
    412
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    575
    +0.11
    0.14
    288
    +0.10
    0.15
    679
    +0.10
    0.13
    Negative Logits
     compromises
    -0.74
     crops
    -0.68
    ious
    -0.67
    xual
    -0.66
    phal
    -0.66
     warrants
    -0.65
     balloons
    -0.65
    itious
    -0.64
    à¨
    -0.64
     disputes
    -0.64
    POSITIVE LOGITS
    were
    0.89
     ARE
    0.83
    're
    0.80
     are
    0.79
    ovie
    0.77
     were
    0.74
    are
    0.74
    ortium
    0.73
    APTER
    0.72
    heet
    0.70
    Activations Density 2.143%

    No Known Activations