© 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. 6523
    Prev
    Next
    INDEX
    Explanations

    references to concepts related to positions, roles, and organizational structures

    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
    58
    +0.13
    0.6%
    533
    +0.10
    0.4%
    110
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.13
    0.19
    40
    +0.10
    0.10
    575
    +0.09
    0.12
    Negative Logits
    ificent
    -0.65
     Rescue
    -0.61
    veland
    -0.61
    EngineDebug
    -0.58
     televised
    -0.58
     unfold
    -0.58
     largeDownload
    -0.57
    ifted
    -0.57
    ilant
    -0.57
    lished
    -0.56
    POSITIVE LOGITS
     altogether
    1.68
     entirely
    1.15
     inhib
    1.05
     prematurely
    0.93
     pesky
    0.88
     barriers
    0.86
     veil
    0.86
     imped
    0.85
     stigma
    0.82
     obsolete
    0.82
    Activations Density 2.819%

    No Known Activations