© 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
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 8-RES_POST_32K-OAI
    6. 4639
    Prev
    Next
    INDEX
    Explanations

    concepts and discussions surrounding truth and belief systems

    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
    295
    +0.10
    0.4%
    14
    +0.10
    0.4%
    691
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.10
    0.19
    691
    +0.10
    0.10
    295
    +0.09
    0.09
    Negative Logits
    onga
    -0.77
    onto
    -0.70
    currently
    -0.69
    ilus
    -0.62
    ourke
    -0.60
    requires
    -0.60
    ems
    -0.59
    animate
    -0.59
     currently
    -0.57
    arget
    -0.57
    POSITIVE LOGITS
     earlier
    1.19
     yesterday
    1.16
     initially
    1.11
     last
    1.08
     originally
    1.03
     beforehand
    0.99
     prior
    0.86
     during
    0.81
    last
    0.77
     previous
    0.76
    Activations Density 4.041%

    No Known Activations