© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 10-RES_POST_32K-OAI
    6. 29477
    Prev
    Next
    INDEX
    Explanations

    symbols and special characters, indicating interest in non-verbal content such as links or formatting

    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
    160
    +0.13
    0.6%
    373
    +0.12
    0.5%
    764
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    530
    +0.13
    0.01
    674
    +0.12
    0.01
    365
    +0.10
    -0.00
    Negative Logits
     Roses
    -0.69
     Aires
    -0.66
     Mahjong
    -0.64
     Standing
    -0.63
    olkien
    -0.60
     Commando
    -0.59
    DonaldTrump
    -0.59
     Walking
    -0.58
    anyahu
    -0.58
     Gingrich
    -0.58
    POSITIVE LOGITS
    usp
    0.91
    ved
    0.91
    psc
    0.89
    sr
    0.83
    btn
    0.81
    hold
    0.79
    location
    0.78
    offer
    0.78
    metadata
    0.77
    scope
    0.77
    Activations Density 0.101%

    No Known Activations