© 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. 6303
    Prev
    Next
    INDEX
    Explanations

    specific names or identifiers, likely personal or character-based terms

    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
    169
    +0.12
    0.5%
    179
    +0.11
    0.5%
    506
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.12
    0.05
    404
    +0.11
    0.05
    556
    +0.10
    0.04
    Negative Logits
    phrine
    -0.73
    ibles
    -0.65
    anmar
    -0.64
    ullivan
    -0.63
    issance
    -0.63
     Pradesh
    -0.62
    ersen
    -0.61
    nz
    -0.59
    tis
    -0.58
    tons
    -0.57
    POSITIVE LOGITS
    unct
    0.78
    ooth
    0.73
    elli
    0.72
    oise
    0.72
    ody
    0.71
    adesh
    0.69
     Reader
    0.67
    ï¸ı
    0.67
    Unix
    0.65
    ati
    0.65
    Activations Density 0.569%

    No Known Activations