© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 5823
    Prev
    Next
    INDEX
    Explanations

    references to essential concepts or ideas within texts

    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
    634
    +0.12
    0.6%
    127
    +0.11
    0.5%
    555
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.12
    0.02
    659
    +0.11
    0.01
    634
    +0.11
    0.01
    Negative Logits
     sight
    -0.74
    bugs
    -0.63
     exagger
    -0.62
    faced
    -0.61
     negligence
    -0.61
     seals
    -0.61
    eur
    -0.61
     bells
    -0.60
    ategory
    -0.59
     slaves
    -0.58
    POSITIVE LOGITS
    entials
    1.41
    endon
    1.37
    ential
    1.15
    entially
    1.11
    ences
    1.06
    endi
    0.97
    linger
    0.96
    enza
    0.95
    inite
    0.95
    enes
    0.93
    Activations Density 0.090%

    No Known Activations