© 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. 4-RES-SM
    11. 10487
    Prev
    Next
    INDEX
    Explanations

    phrases that involve requests or polite commands

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    307
    +0.16
    0.9%
    25
    +0.15
    0.8%
    358
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    72
    +0.16
    0.07
    476
    +0.15
    0.08
    472
    +0.11
    0.06
    Negative Logits
    ¾
    -4.91
    ij
    -4.83
    Ħ
    -4.64
    ĻĤ
    -4.61
    Ĥ¬
    -4.58
    ŀ
    -4.54
    Ģ
    -4.52
    ļ
    -4.47
    ĥ
    -4.45
    ·¸
    -4.37
    POSITIVE LOGITS
     please
    3.24
     you
    2.82
    please
    2.74
     consider
    2.56
    you
    2.34
     let
    2.23
    your
    2.09
     visit
    2.08
     listen
    2.05
     check
    2.04
    Activations Density 0.752%

    No Known Activations