© 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. 31836
    Prev
    Next
    INDEX
    Explanations

    references to the word "ick" and its variations

    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
    285
    +0.11
    0.5%
    276
    +0.11
    0.5%
    683
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.11
    0.03
    232
    +0.11
    0.02
    683
    +0.11
    0.02
    Negative Logits
    ONT
    -0.67
     ILCS
    -0.67
    orescence
    -0.66
    PsyNetMessage
    -0.63
    ãģ®å
    -0.62
    BIP
    -0.60
    uthor
    -0.60
    ãĤ£
    -0.60
    ATIONAL
    -0.58
     Mazda
    -0.58
    POSITIVE LOGITS
    leness
    1.26
    ety
    1.22
    ering
    1.09
    ett
    1.09
    eters
    1.08
    eter
    1.05
    etts
    1.03
    lers
    0.98
    erman
    0.98
    ery
    0.98
    Activations Density 0.177%

    No Known Activations