© 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. 8331
    Prev
    Next
    INDEX
    Explanations

    instances and references to different string types and data structures in programming

    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
    61
    +0.11
    0.5%
    399
    +0.10
    0.5%
    217
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    158
    +0.11
    0.03
    326
    +0.10
    0.02
    647
    +0.10
    0.02
    Negative Logits
    laus
    -0.83
    gaard
    -0.82
    cart
    -0.80
    alos
    -0.79
    heimer
    -0.77
    arak
    -0.76
    ebus
    -0.76
    iaries
    -0.74
    ilver
    -0.73
    kens
    -0.72
    POSITIVE LOGITS
     liter
    0.99
     literal
    0.91
     representation
    0.90
     representations
    0.83
    ãĤ©
    0.79
    ify
    0.76
     expression
    0.76
    izer
    0.74
     readable
    0.73
     encoded
    0.73
    Activations Density 0.092%

    No Known Activations