© 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. Residual Post - 32k
    5. 10-RES_POST_32K-OAI
    6. 29499
    Prev
    Next
    INDEX
    Explanations

    structural elements and formatting indicators in text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_10.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.10.hook_resid_post
    Hook Layer
    10
    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
    295
    +0.13
    0.6%
    182
    +0.12
    0.5%
    530
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.13
    0.03
    530
    +0.12
    0.03
    326
    +0.10
    0.05
    Negative Logits
     Wen
    -0.71
    arine
    -0.70
    pad
    -0.69
    gdala
    -0.69
    iband
    -0.68
    haired
    -0.67
     Cou
    -0.67
     Camel
    -0.67
    Pad
    -0.66
     thighs
    -0.66
    POSITIVE LOGITS
    INFO
    1.87
    Info
    1.66
    info
    1.66
     Info
    1.60
     info
    1.37
     INFO
    1.30
    WARN
    1.30
     Information
    1.08
    ERROR
    1.07
    WARNING
    1.06
    Activations Density 0.325%

    No Known Activations