© 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. 28089
    Prev
    Next
    INDEX
    Explanations

    expressions of the word "give" in various contexts

    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
    13
    +0.11
    0.5%
    238
    +0.10
    0.4%
    426
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    279
    +0.11
    0.06
    426
    +0.10
    0.03
    230
    +0.10
    0.04
    Negative Logits
    sole
    -0.70
    thodox
    -0.69
    hement
    -0.68
    TERN
    -0.67
    ntil
    -0.62
    GROUND
    -0.62
    ruction
    -0.60
    utor
    -0.60
    ritch
    -0.59
    ãĥIJ
    -0.58
    POSITIVE LOGITS
     slightest
    0.80
     fools
    0.73
     knots
    0.72
    hin
    0.72
     angles
    0.70
    enance
    0.68
    atform
    0.67
    gets
    0.67
     hesitate
    0.66
     hesitation
    0.65
    Activations Density 0.363%

    No Known Activations