© 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. 6884
    Prev
    Next
    INDEX
    Explanations

    terms related to operational processes and management

    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
    756
    +0.14
    0.6%
    288
    +0.13
    0.6%
    540
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    540
    +0.14
    0.04
    288
    +0.13
    0.03
    216
    +0.13
    0.02
    Negative Logits
    hus
    -0.71
    udeau
    -0.69
    cil
    -0.67
    sonian
    -0.65
    apple
    -0.64
    yu
    -0.63
    ãĤ©
    -0.63
    ioxide
    -0.62
    bitious
    -0.61
    vana
    -0.61
    POSITIVE LOGITS
    ally
    1.12
    ality
    0.91
    ional
    0.89
    eering
    0.87
    ament
    0.80
     manuals
    0.78
    ername
    0.74
     tempo
    0.73
     manual
    0.70
    eers
    0.69
    Activations Density 0.120%

    No Known Activations