© 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. 1-RES-SM
    11. 28299
    Prev
    Next
    INDEX
    Explanations

    words that indicate general agreement or addition in a context

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    23
    +0.19
    1.1%
    14
    +0.12
    0.7%
    78
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    377
    +0.19
    0.09
    78
    +0.12
    0.08
    14
    +0.12
    0.07
    Negative Logits
    ħ
    -2.07
    ±
    -2.04
    Ģ
    -1.95
    Ħ
    -1.94
    ĨĴ
    -1.85
    IJ
    -1.83
    ļ
    -1.78
    Ŀ
    -1.76
    ³
    -1.73
    ¢
    -1.72
    POSITIVE LOGITS
    ffected
    1.66
     subdu
    1.46
     invaded
    1.46
     metast
    1.41
    ++){
    1.40
     deprec
    1.38
    ?_
    1.37
     mild
    1.36
    comings
    1.35
     diffus
    1.35
    Activations Density 0.748%

    No Known Activations