© 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. 6139
    Prev
    Next
    INDEX
    Explanations

    the term "known" in various contexts

    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
    373
    +0.15
    0.7%
    689
    +0.14
    0.7%
    94
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    383
    +0.15
    0.04
    689
    +0.14
    0.03
    42
    +0.11
    0.03
    Negative Logits
     fixation
    -0.64
    oire
    -0.61
    ï¸ı
    -0.58
    poral
    -0.56
    sense
    -0.56
    ateurs
    -0.54
    eur
    -0.54
    olicy
    -0.52
     acquisition
    -0.52
    pled
    -0.51
    POSITIVE LOGITS
     collectively
    0.77
    amed
    0.70
     termed
    0.67
     Known
    0.67
     abbrevi
    0.66
     euphem
    0.65
     as
    0.65
    onym
    0.64
     known
    0.63
     commonly
    0.63
    Activations Density 0.145%

    No Known Activations