© 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 EvalsExportsinterp-engineNEWAPI 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. 7960
    Prev
    Next
    INDEX
    Explanations

    references to political members, parties, and legislative terms

    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
    373
    +0.14
    0.6%
    712
    +0.11
    0.5%
    689
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    731
    +0.14
    0.03
    477
    +0.11
    0.03
    708
    +0.09
    0.04
    Negative Logits
    mberg
    -0.81
     listens
    -0.75
    otin
    -0.74
    qv
    -0.71
    lets
    -0.71
    tons
    -0.69
    qus
    -0.69
     listened
    -0.69
    ktop
    -0.68
    ctrl
    -0.67
    POSITIVE LOGITS
    ¥ŀ
    0.79
     None
    0.73
     Died
    0.63
     ENG
    0.62
     Moderate
    0.59
     Partial
    0.59
     Female
    0.59
     Reported
    0.59
     Ùħ
    0.58
    Legendary
    0.58
    Activations Density 0.392%

    No Known Activations