© 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 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. 7134
    Prev
    Next
    INDEX
    Explanations

    references to divisions or segments within a whole

    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
    218
    +0.13
    0.6%
    190
    +0.12
    0.6%
    148
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    344
    +0.13
    0.04
    190
    +0.12
    0.04
    597
    +0.10
    0.04
    Negative Logits
     Liberation
    -0.61
     TOP
    -0.61
     diction
    -0.59
     Pax
    -0.58
    rament
    -0.57
     Colors
    -0.56
     Tackle
    -0.55
     Ribbon
    -0.55
     Courier
    -0.55
     CoC
    -0.54
    POSITIVE LOGITS
    way
    0.76
    irez
    0.73
    heartedly
    0.72
    SpaceEngineers
    0.72
    iece
    0.70
    eeper
    0.70
    nown
    0.69
    fell
    0.69
     thereof
    0.68
     ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³
    0.68
    Activations Density 0.166%

    No Known Activations