© 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. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 2-RES-SM
    11. 10714
    Prev
    Next
    INDEX
    Explanations

    words related to legal enforcement and violations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__res-sm_processed/2-res-sm
    Prompts (Dashboard)
    32,768 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.2.hook_resid_post
    Hook Layer
    2
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    419
    +0.21
    1.2%
    17
    +0.17
    0.9%
    231
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    419
    +0.21
    0.05
    495
    +0.17
    0.05
    18
    +0.14
    0.04
    Negative Logits
    ĸ
    -2.34
    Ĩ
    -2.21
    Ĺ
    -2.19
    §
    -2.11
    ¼
    -1.99
    ª
    -1.94
    ľĵ
    -1.93
    ¯
    -1.92
    ¸
    -1.92
    º
    -1.92
    POSITIVE LOGITS
    ifiable
    1.74
    ized
    1.63
    anted
    1.58
     nor
    1.57
     recess
    1.40
     dire
    1.37
    ifying
    1.36
     indict
    1.31
    alled
    1.30
    sure
    1.29
    Activations Density 0.386%

    No Known Activations