© 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. 8-RES_POST_32K-OAI
    11. 7290
    Prev
    Next
    INDEX
    Explanations

    the word "somewhat" 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
    649
    +0.11
    0.5%
    618
    +0.10
    0.5%
    195
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    151
    +0.11
    0.03
    101
    +0.10
    0.03
    331
    +0.09
    0.03
    Negative Logits
    itures
    -0.78
    mson
    -0.74
    ULTS
    -0.73
    yer
    -0.73
    yes
    -0.72
    abad
    -0.72
    XL
    -0.72
    ulkan
    -0.71
    quality
    -0.71
    ivable
    -0.71
    POSITIVE LOGITS
     misleading
    0.79
     resembling
    0.77
     obscured
    0.77
     overlapping
    0.76
     obscure
    0.74
     surprising
    0.74
     neb
    0.72
     confusing
    0.72
     ironically
    0.72
     controvers
    0.71
    Activations Density 0.045%

    No Known Activations