© 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. 6015
    Prev
    Next
    INDEX
    Explanations

    phrases and concepts related to leaving an impact or trace on something

    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
    459
    +0.12
    0.5%
    175
    +0.12
    0.5%
    506
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    36
    +0.12
    0.03
    636
    +0.12
    0.02
    402
    +0.10
    0.03
    Negative Logits
    Cro
    -0.68
    hy
    -0.67
    theless
    -0.66
    uits
    -0.64
    soever
    -0.64
    hedon
    -0.62
    itsch
    -0.61
     horizont
    -0.59
    soon
    -0.58
    scl
    -0.58
    POSITIVE LOGITS
     lasting
    1.00
     trail
    0.87
     dent
    0.83
     impression
    0.82
     footprints
    0.79
     imprint
    0.78
     footprint
    0.77
     mark
    0.74
    ãĥĩãĤ£
    0.74
    ullivan
    0.73
    Activations Density 0.191%

    No Known Activations