© 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 EvalsExportsAPI 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. 30132
    Prev
    Next
    INDEX
    Explanations

    references to military movements and actions

    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
    430
    +0.11
    0.5%
    34
    +0.11
    0.5%
    666
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    413
    +0.11
    0.03
    196
    +0.11
    0.03
    305
    +0.11
    0.04
    Negative Logits
    ï
    -0.78
    opia
    -0.75
     millennia
    -0.75
    !?"
    -0.73
    éŃĶ
    -0.71
    limits
    -0.71
    âĶĢâĶĢ
    -0.71
     Machina
    -0.70
     creators
    -0.68
    opian
    -0.68
    POSITIVE LOGITS
     Prayer
    1.12
     Evening
    1.04
     Yesterday
    0.90
    noon
    0.90
     Horses
    0.88
     Meeting
    0.87
     rece
    0.87
     Called
    0.84
     Dept
    0.84
     hrs
    0.81
    Activations Density 0.728%

    No Known Activations