© 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
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 10-RES_POST_32K-OAI
    6. 27154
    Prev
    Next
    INDEX
    Explanations

    instances of the word "change" and its variations, indicating a focus on transformations or shifts in situations

    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
    756
    +0.11
    0.5%
    438
    +0.11
    0.5%
    109
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    271
    +0.11
    0.05
    438
    +0.11
    0.04
    750
    +0.10
    0.04
    Negative Logits
    panic
    -0.72
    isy
    -0.66
    Adds
    -0.64
    bey
    -0.63
    leted
    -0.61
    warning
    -0.61
    iar
    -0.60
    rying
    -0.59
    shots
    -0.58
     complains
    -0.58
    POSITIVE LOGITS
     radically
    1.04
     drastically
    1.03
     dramatically
    1.01
     fundamentally
    0.97
     profoundly
    0.95
     markedly
    0.86
     minds
    0.86
     fortunes
    0.84
     hands
    0.81
     attitudes
    0.80
    Activations Density 0.401%

    No Known Activations