© 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. 28595
    Prev
    Next
    INDEX
    Explanations

    discussions centered around consensus opinions and prevailing views among experts or groups

    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
    21
    +0.14
    0.6%
    702
    +0.12
    0.5%
    401
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    679
    +0.14
    0.12
    401
    +0.12
    0.10
    21
    +0.09
    0.07
    Negative Logits
    viol
    -0.70
    quished
    -0.69
    teness
    -0.67
    rir
    -0.66
    ILE
    -0.66
    interrupted
    -0.65
    \/\/
    -0.65
     cumbers
    -0.62
    NULL
    -0.62
    ulence
    -0.61
    POSITIVE LOGITS
     consensus
    1.05
     prevailing
    0.98
     prevailed
    0.97
     assumption
    0.89
     wisdom
    0.89
     favoring
    0.89
     view
    0.88
    ensus
    0.83
     prev
    0.82
     among
    0.80
    Activations Density 1.610%

    No Known Activations