© 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. 10-RES_POST_32K-OAI
    11. 29858
    Prev
    Next
    INDEX
    Explanations

    phrases related to recovery and progress

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_10.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.10.hook_resid_post
    Hook Layer
    10
    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
    373
    +0.10
    0.4%
    425
    +0.10
    0.4%
    46
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    113
    +0.10
    0.05
    526
    +0.10
    -0.01
    242
    +0.09
    0.05
    Negative Logits
     unavailable
    -0.73
     Reserved
    -0.73
     scorn
    -0.71
     deprived
    -0.69
     Argument
    -0.69
    cens
    -0.69
     insulted
    -0.68
    icides
    -0.67
    ERROR
    -0.67
    anamo
    -0.66
    POSITIVE LOGITS
     groove
    0.99
     track
    0.97
     pace
    0.93
     rhythm
    0.91
     shape
    0.90
     sustainable
    0.89
    shape
    0.85
    Track
    0.85
    Course
    0.82
     footing
    0.81
    Activations Density 1.453%

    No Known Activations