Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 8-RES_POST_32K-OAI
    6. 5138
    Prev
    Next
    INDEX
    Explanations

    geographic locations and associated places

    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
    445
    +0.10
    0.4%
    658
    +0.09
    0.4%
    530
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.10
    0.19
    575
    +0.09
    0.14
    677
    +0.09
    0.09
    Negative Logits
    athi
    -0.71
     @@
    -0.65
     resemblance
    -0.62
     surveyed
    -0.60
     Coup
    -0.58
     thumbs
    -0.56
     handwriting
    -0.56
     polled
    -0.55
     Amendments
    -0.55
    iqueness
    -0.54
    POSITIVE LOGITS
    livion
    1.05
     destinations
    0.98
    venge
    0.89
     whence
    0.83
     destination
    0.81
    orage
    0.81
     fray
    0.79
     shores
    0.79
     coffers
    0.78
    safe
    0.74
    Activations Density 2.972%

    No Known Activations