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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 5416
    Prev
    Next
    INDEX
    Explanations

    references to experiments and studies

    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
    746
    +0.11
    0.5%
    497
    +0.10
    0.5%
    78
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    239
    +0.11
    0.04
    677
    +0.10
    0.04
    665
    +0.09
    0.04
    Negative Logits
    ulkan
    -0.71
    stick
    -0.69
    loaded
    -0.68
    ipedia
    -0.67
    dra
    -0.65
    iw
    -0.65
    ifted
    -0.65
     Leilan
    -0.65
    icator
    -0.65
    Broad
    -0.64
    POSITIVE LOGITS
     jointly
    0.83
    uate
    0.81
     Lans
    0.81
    pez
    0.73
     concurrently
    0.70
     feasibility
    0.69
     collabor
    0.64
     annually
    0.63
     conducted
    0.62
    udo
    0.61
    Activations Density 0.149%

    No Known Activations