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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    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. 5020
    Prev
    Next
    INDEX
    Explanations

    instances of the word "escape" and its variations

    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
    100
    +0.12
    0.5%
    728
    +0.10
    0.5%
    288
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    100
    +0.12
    0.02
    207
    +0.10
    0.02
    516
    +0.10
    0.02
    Negative Logits
    eous
    -0.70
    sonian
    -0.69
    matically
    -0.66
    iop
    -0.62
    nan
    -0.61
    ificent
    -0.60
     stead
    -0.59
     jo
    -0.57
    wart
    -0.57
    shed
    -0.57
    POSITIVE LOGITS
     hatch
    0.95
     velocity
    0.88
     Velocity
    0.83
    APE
    0.83
    HTML
    0.80
    ibility
    0.77
     routes
    0.76
    ously
    0.76
     Torment
    0.74
    vable
    0.72
    Activations Density 0.061%

    No Known Activations