© 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. 5629
    Prev
    Next
    INDEX
    Explanations

    the word "where" in various contexts

    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
    236
    +0.12
    0.5%
    296
    +0.11
    0.5%
    373
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    236
    +0.12
    0.04
    186
    +0.11
    0.03
    300
    +0.09
    0.03
    Negative Logits
    bart
    -0.70
    UX
    -0.69
    ded
    -0.69
    ature
    -0.67
    esis
    -0.67
    MER
    -0.66
    ude
    -0.66
    eme
    -0.66
    edy
    -0.65
    tm
    -0.63
    POSITIVE LOGITS
     exactly
    0.85
     else
    0.82
    abouts
    0.81
     nearest
    0.65
     Rivers
    0.59
     Else
    0.58
     Dat
    0.58
     Chero
    0.58
     roads
    0.57
    rils
    0.56
    Activations Density 0.061%

    No Known Activations