© 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
    3. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 8-RES_POST_32K-OAI
    11. 8485
    Prev
    Next
    INDEX
    Explanations

    terms related to research and its various aspects

    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
    362
    +0.14
    0.6%
    314
    +0.13
    0.6%
    462
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.14
    0.12
    575
    +0.13
    0.09
    41
    +0.10
    0.06
    Negative Logits
    lees
    -0.73
    .</
    -0.73
    ernels
    -0.66
    liners
    -0.65
    .'
    -0.63
    .ãĢį
    -0.62
    !.
    -0.62
    !:
    -0.61
    els
    -0.61
    !'
    -0.61
    POSITIVE LOGITS
     equation
    0.81
     standpoint
    0.79
    prise
    0.78
     regimen
    0.73
     solution
    0.73
     nightmare
    0.71
     headache
    0.71
     avenue
    0.69
     mindset
    0.69
     source
    0.69
    Activations Density 0.906%

    No Known Activations