© 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. 28521
    Prev
    Next
    INDEX
    Explanations

    instances of the word "has" and related phrases indicating possession or actions taken

    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
    718
    +0.12
    0.6%
    86
    +0.12
    0.6%
    690
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    668
    +0.12
    0.12
    344
    +0.12
    0.12
    7
    +0.10
    0.10
    Negative Logits
     encounters
    -0.84
     sightings
    -0.78
     crashes
    -0.69
    Discover
    -0.68
    photos
    -0.68
     highs
    -0.67
     ghosts
    -0.67
     Pyth
    -0.66
     mates
    -0.65
     encountering
    -0.65
    POSITIVE LOGITS
     unequivocally
    0.90
     wisely
    0.88
     stated
    0.88
     acknowledged
    0.86
     signaled
    0.86
     reiterated
    0.85
     indicated
    0.85
     affirmed
    0.85
     nominated
    0.83
     rightly
    0.83
    Activations Density 0.816%

    No Known Activations