© 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. 4215
    Prev
    Next
    INDEX
    Explanations

    the word 'even' in various contexts

    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%
    37
    +0.10
    0.5%
    219
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    133
    +0.14
    0.05
    243
    +0.10
    0.04
    660
    +0.09
    0.04
    Negative Logits
    urst
    -0.94
    kefeller
    -0.72
    definition
    -0.71
    uti
    -0.71
    roots
    -0.70
    serv
    -0.69
    hered
    -0.67
    broad
    -0.66
    ute
    -0.66
    qi
    -0.66
    POSITIVE LOGITS
     haha
    0.74
     worse
    0.73
     XD
    0.71
     though
    0.71
     tho
    0.70
     funn
    0.69
     ones
    0.65
     better
    0.65
     compliment
    0.64
     anyways
    0.60
    Activations Density 0.171%

    No Known Activations