© 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. 29074
    Prev
    Next
    INDEX
    Explanations

    references to single entities or units

    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
    251
    +0.11
    0.5%
    238
    +0.10
    0.5%
    359
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    36
    +0.11
    0.12
    328
    +0.10
    0.07
    250
    +0.09
    0.07
    Negative Logits
     spoiler
    -0.64
     wiser
    -0.63
     spoilers
    -0.62
     safer
    -0.61
     fortun
    -0.61
     superflu
    -0.61
     reinforcing
    -0.59
     scapego
    -0.58
     derog
    -0.58
     respects
    -0.57
    POSITIVE LOGITS
     single
    1.03
     typical
    1.02
     given
    0.96
     average
    0.95
    average
    0.90
    each
    0.84
     total
    0.84
    verages
    0.82
    ©¶æ
    0.80
     maximum
    0.80
    Activations Density 0.737%

    No Known Activations