© 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. 10-RES_POST_32K-OAI
    11. 31897
    Prev
    Next
    INDEX
    Explanations

    instances of communication and formal interactions

    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
    67
    +0.12
    0.5%
    390
    +0.11
    0.5%
    275
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.12
    0.11
    67
    +0.11
    0.08
    59
    +0.10
    0.06
    Negative Logits
    something
    -0.73
    artifacts
    -0.71
    ortality
    -0.71
    void
    -0.69
     sorcery
    -0.68
    oneliness
    -0.68
    Untitled
    -0.68
    ãĥĻ
    -0.66
    stones
    -0.65
     madness
    -0.64
    POSITIVE LOGITS
     stakeholders
    1.27
     affected
    1.06
     customers
    0.97
     applicants
    0.97
     clients
    0.97
     suppliers
    0.96
     claimants
    0.96
     respondents
    0.95
     landowners
    0.94
     vendors
    0.93
    Activations Density 1.041%

    No Known Activations