© 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. 8-RES_POST_32K-OAI
    11. 6805
    Prev
    Next
    INDEX
    Explanations

    unique character sequences or symbols

    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
    +0.20
    0.9%
    92
    +0.09
    0.4%
    218
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    675
    +0.20
    0.26
    107
    +0.09
    0.22
    48
    +0.09
    0.19
    Negative Logits
     displacement
    -0.81
     disadvantage
    -0.77
     atten
    -0.76
     dil
    -0.73
     differential
    -0.73
     retained
    -0.72
     anomal
    -0.72
     polymorph
    -0.72
     psi
    -0.72
     duplication
    -0.72
    POSITIVE LOGITS
    ðŁĺ
    1.82
    ðŁ
    1.80
    ðŁij
    1.52
    âĿ
    1.40
    ï¸ı
    1.34
     ðŁ
    1.28
     ðŁij
    1.24
    âľ
    1.24
     âĿ
    1.18
     ðŁĺ
    1.18
    Activations Density 2.534%

    No Known Activations