© 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. 6376
    Prev
    Next
    INDEX
    Explanations

    numeric identifiers related to digital files or events

    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
    88
    +0.13
    0.6%
    695
    +0.13
    0.6%
    373
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    23
    +0.13
    0.05
    107
    +0.13
    0.05
    254
    +0.10
    0.04
    Negative Logits
     Boyd
    -0.74
    wire
    -0.68
    hands
    -0.68
    und
    -0.68
    20439
    -0.67
    laws
    -0.66
    undy
    -0.65
    anchester
    -0.64
    oft
    -0.64
    light
    -0.63
    POSITIVE LOGITS
     ->
    2.15
    ->
    1.98
     âĨĴ
    1.26
    âĨĴ
    1.10
    =>
    1.07
    -->
    1.06
     -->
    1.03
    Į
    1.02
    ::
    0.98
     =>
    0.97
    Activations Density 0.371%

    No Known Activations