© 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. 32317
    Prev
    Next
    INDEX
    Explanations

    specific numerical or coded values associated with data structures and programming functions

    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
    307
    +0.12
    0.5%
    398
    +0.11
    0.5%
    753
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    36
    +0.12
    0.03
    476
    +0.11
    0.02
    430
    +0.10
    0.03
    Negative Logits
    WAYS
    -0.80
     arrang
    -0.78
     virtues
    -0.70
     incent
    -0.70
     idols
    -0.69
     compromises
    -0.69
     designs
    -0.68
    Ul
    -0.65
     weaknesses
    -0.65
     lifestyles
    -0.64
    POSITIVE LOGITS
     GOODMAN
    0.95
    udeb
    0.81
    itialized
    0.77
    eree
    0.77
     repro
    0.76
     shader
    0.76
     PCR
    0.74
    FML
    0.73
     "$:/
    0.73
     Reloaded
    0.73
    Activations Density 0.074%

    No Known Activations