© 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. 6561
    Prev
    Next
    INDEX
    Explanations

    phrases indicating relationships or connections, especially those involving temporal references

    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
    +0.11
    0.5%
    218
    +0.11
    0.5%
    448
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    305
    +0.11
    0.04
    269
    +0.11
    0.03
    545
    +0.11
    0.03
    Negative Logits
    Planet
    -0.74
    Enhanced
    -0.69
     Tradable
    -0.69
    Dynamic
    -0.63
     inferior
    -0.62
     superior
    -0.62
    Excellent
    -0.62
    Ability
    -0.61
    Compan
    -0.61
    USB
    -0.59
    POSITIVE LOGITS
     1945
    0.86
     2005
    0.83
     1918
    0.83
     1995
    0.83
     2010
    0.80
     2003
    0.78
     2006
    0.78
     2007
    0.78
     1899
    0.78
     1985
    0.78
    Activations Density 0.147%

    No Known Activations