© 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. 10-RES_POST_32K-OAI
    11. 28354
    Prev
    Next
    INDEX
    Explanations

    instances of the word "run" and its variations

    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
    351
    +0.11
    0.5%
    715
    +0.11
    0.5%
    290
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    351
    +0.11
    0.04
    70
    +0.11
    0.04
    764
    +0.09
    0.03
    Negative Logits
    uctor
    -0.70
    respect
    -0.68
     arrang
    -0.63
     recomm
    -0.62
    ciples
    -0.61
     conduc
    -0.60
    emate
    -0.59
    ashington
    -0.58
     unfold
    -0.58
    lav
    -0.56
    POSITIVE LOGITS
     into
    1.04
     away
    0.96
    awed
    0.83
     INTO
    0.83
     screaming
    0.77
     af
    0.76
     awa
    0.75
     hither
    0.75
     smack
    0.74
     onto
    0.74
    Activations Density 0.182%

    No Known Activations