© 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. 6747
    Prev
    Next
    INDEX
    Explanations

    iterations of the word "plan" and its derivatives

    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
    665
    +0.13
    0.6%
    239
    +0.12
    0.5%
    602
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    665
    +0.13
    0.05
    723
    +0.12
    0.03
    641
    +0.10
    0.03
    Negative Logits
    bery
    -0.67
    ublic
    -0.67
    rants
    -0.65
    govtrack
    -0.65
    aghd
    -0.65
    bour
    -0.63
    ixels
    -0.61
    mint
    -0.60
    pires
    -0.59
    è£ħ
    -0.58
    POSITIVE LOGITS
     ahead
    0.95
     accordingly
    0.88
     beforehand
    0.84
     Parenthood
    0.80
     vacations
    0.79
    paren
    0.73
    dule
    0.73
    tera
    0.71
    acci
    0.69
     aloud
    0.69
    Activations Density 0.162%

    No Known Activations