© 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. 5361
    Prev
    Next
    INDEX
    Explanations

    specific token patterns, particularly those involving special characters rather than standard text

    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
    373
    +0.16
    0.8%
    426
    +0.11
    0.5%
    514
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    759
    +0.16
    0.27
    514
    +0.11
    0.25
    740
    +0.10
    0.24
    Negative Logits
    ipop
    -0.68
     Izan
    -0.67
     Jed
    -0.61
     Emblem
    -0.59
    agos
    -0.59
     +++
    -0.58
     Alto
    -0.58
     Samar
    -0.58
    WARD
    -0.58
     Crush
    -0.57
    POSITIVE LOGITS
    s
    0.91
    \'
    0.84
    tis
    0.82
    £
    0.81
    ves
    0.80
    ¢
    0.78
    felt
    0.77
    âĤ¬
    0.77
    ll
    0.77
    ÃŃs
    0.76
    Activations Density 2.495%

    No Known Activations