© 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. 10066
    Prev
    Next
    INDEX
    Explanations

    phrases that refer to historical events or noteworthy timelines

    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
    274
    +0.23
    1.3%
    156
    +0.18
    1.0%
    291
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    291
    +0.23
    0.07
    145
    +0.18
    0.08
    137
    +0.14
    0.08
    Negative Logits
    ģ
    -3.10
    ŀ
    -2.52
    ¢
    -2.46
    ī
    -2.38
    ³
    -2.36
    Ł
    -2.35
    ł
    -2.33
    Īĺ
    -2.33
    ¤
    -2.32
    ¥
    -2.32
    POSITIVE LOGITS
     standpoint
    3.01
     outset
    2.72
     earlier
    2.39
     previous
    2.09
     inception
    2.08
     outside
    2.07
     archives
    1.97
     perspective
    1.95
     earliest
    1.89
     periphery
    1.83
    Activations Density 0.765%

    No Known Activations