© 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 EvalsExportsinterp-engineNEWAPI 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. 28062
    Prev
    Next
    INDEX
    Explanations

    mathematical expressions and relationships among values or variables

    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
    82
    +0.11
    0.5%
    702
    +0.10
    0.5%
    720
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    720
    +0.11
    0.07
    702
    +0.10
    0.06
    369
    +0.10
    0.03
    Negative Logits
    76561
    -0.76
    ork
    -0.68
    URA
    -0.66
     Flavoring
    -0.66
    ITNESS
    -0.63
     garg
    -0.62
    aph
    -0.62
    Ult
    -0.60
     Traps
    -0.60
    gel
    -0.59
    POSITIVE LOGITS
     Right
    2.00
     Left
    2.00
     left
    1.99
    Left
    1.98
    Right
    1.93
     RIGHT
    1.92
     right
    1.90
    left
    1.89
    right
    1.85
     leftists
    1.33
    Activations Density 2.210%

    No Known Activations