© 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. 31852
    Prev
    Next
    INDEX
    Explanations

    contexts related to notable events or significant occurrences

    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
    481
    +0.24
    2.6%
    447
    +0.21
    2.3%
    480
    +0.21
    2.3%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    481
    +0.24
    0.68
    480
    +0.21
    0.26
    266
    +0.21
    0.67
    Negative Logits
     somewhere
    -0.31
     halfway
    -0.28
     somew
    -0.27
     sometime
    -0.26
     elsewhere
    -0.26
     alternatively
    -0.25
     due
    -0.25
     maybe
    -0.24
     sooner
    -0.24
     incidentally
    -0.24
    POSITIVE LOGITS
    inventoryQuantity
    0.42
     teasp
    0.42
    ĸļ士
    0.42
    Ę
    0.41
    Ă
    0.41
    Č
    0.41
    ü
    0.41
    ġ
    0.41
    ø
    0.41
    ć
    0.41
    Activations Density 56.311%

    No Known Activations