© 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. 31200
    Prev
    Next
    INDEX
    Explanations

    the preposition "in"

    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
    314
    +0.16
    0.8%
    637
    +0.13
    0.6%
    616
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    616
    +0.16
    0.04
    123
    +0.13
    0.04
    22
    +0.10
    0.03
    Negative Logits
     OMG
    -0.67
     morale
    -0.64
     grate
    -0.62
     McGill
    -0.61
     DOI
    -0.60
     bye
    -0.57
    DragonMagazine
    -0.57
     Brach
    -0.57
     SIG
    -0.56
     shake
    -0.56
    POSITIVE LOGITS
    strument
    1.77
    jury
    1.69
    visible
    1.66
    iti
    1.63
    jured
    1.63
    vasive
    1.60
    cluded
    1.55
    clusive
    1.55
    structed
    1.54
    valid
    1.51
    Activations Density 0.182%

    No Known Activations