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

    Neuronpedia

    Jacobian LensNEW
    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. 29066
    Prev
    Next
    INDEX
    Explanations

    mentions of perception and approval regarding various subjects

    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
    380
    +0.15
    0.7%
    2
    +0.12
    0.5%
    293
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    314
    +0.15
    -0.00
    232
    +0.12
    0.03
    380
    +0.10
    0.01
    Negative Logits
     exagger
    -0.71
     inaccur
    -0.71
     fracturing
    -0.69
     Sith
    -0.66
     eccentric
    -0.63
     irregularities
    -0.62
     acrylic
    -0.62
     abbrevi
    -0.62
     Orthodox
    -0.61
     imperfect
    -0.60
    POSITIVE LOGITS
    iving
    1.52
    ivable
    1.45
    ving
    1.28
    ives
    1.24
    ivers
    1.16
    ved
    1.12
    vable
    1.11
    ve
    1.09
    ves
    1.08
    ipt
    1.07
    Activations Density 0.243%

    No Known Activations