© 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. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 4-RES-SM
    11. 10858
    Prev
    Next
    INDEX
    Explanations

    terms related to drivers and their respective solutions

    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
    156
    +0.41
    2.8%
    348
    +0.29
    2.0%
    466
    +0.24
    1.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    25
    +0.41
    0.13
    156
    +0.29
    0.12
    466
    +0.24
    0.10
    Negative Logits
    ij
    -2.17
    ī
    -2.13
    ĥ½
    -2.08
    Ī
    -2.08
    Ħ
    -1.91
    »
    -1.88
    ·¸
    -1.86
    -1.84
    ↵
    -1.84
    <|outofrange|>
    -1.84
    POSITIVE LOGITS
     respectively
    1.99
    .
    1.80
     alike
    1.66
     theirs
    1.64
    iginally
    1.42
    .).
    1.42
    .
    1.42
    .\
    1.41
     wont
    1.39
    .*;
    1.37
    Activations Density 2.251%

    No Known Activations