© 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. 14381
    Prev
    Next
    INDEX
    Explanations

    the term "such."

    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
    21
    +0.19
    1.3%
    348
    +0.18
    1.2%
    201
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    201
    +0.19
    0.04
    190
    +0.18
    0.04
    180
    +0.10
    0.03
    Negative Logits
    ĭ
    -3.01
    ľ
    -2.88
    ľĵ
    -2.87
    ¹
    -2.85
    ĸ´
    -2.83
    »
    -2.80
    ī
    -2.75
    ·¸
    -2.74
    ¾
    -2.73
    º
    -2.71
    POSITIVE LOGITS
     as
    2.18
    as
    1.36
    ness
    1.30
     luck
    1.30
    oths
    1.30
    zymes
    1.24
    mina
    1.23
     como
    1.21
    iya
    1.20
    hljs
    1.19
    Activations Density 0.067%

    No Known Activations