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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 31757
    Prev
    Next
    INDEX
    Explanations

    references to specific topics or themes in a 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
    195
    +0.18
    0.8%
    655
    +0.10
    0.5%
    737
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    737
    +0.18
    0.03
    252
    +0.10
    0.03
    195
    +0.10
    0.02
    Negative Logits
    ignt
    -0.84
    oÄŁ
    -0.73
    owicz
    -0.70
    xon
    -0.69
    hovah
    -0.67
     corrid
    -0.67
    ypes
    -0.66
    Lv
    -0.65
    axy
    -0.65
    opers
    -0.65
    POSITIVE LOGITS
    matter
    1.06
     matter
    0.93
     topics
    0.93
     topic
    0.92
    ually
    0.86
     area
    0.85
     covered
    0.81
    ivity
    0.79
     discussed
    0.76
    uality
    0.76
    Activations Density 0.089%

    No Known Activations