© 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. 8-RES_POST_32K-OAI
    11. 5895
    Prev
    Next
    INDEX
    Explanations

    punctuation marks that denote the beginning and end of quotes or speech

    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
    +0.19
    0.9%
    125
    +0.09
    0.4%
    420
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    426
    +0.19
    0.06
    352
    +0.09
    0.07
    23
    +0.09
    0.06
    Negative Logits
    İĭ
    -0.81
     microscope
    -0.68
    EngineDebug
    -0.67
    ¥µ
    -0.65
     zoning
    -0.64
     Phant
    -0.62
    asers
    -0.60
     regul
    -0.60
    anke
    -0.57
     ranking
    -0.57
    POSITIVE LOGITS
     +
    0.85
    sic
    0.82
    +(
    0.77
     ++
    0.74
    к
    0.73
    \\\\
    0.72
    fi
    0.72
    word
    0.72
    hov
    0.71
     +=
    0.70
    Activations Density 0.333%

    No Known Activations