© 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. 2-RES-SM
    11. 13212
    Prev
    Next
    INDEX
    Explanations

    references to software programming concepts and configurations

    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/2-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.2.hook_resid_post
    Hook Layer
    2
    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
    320
    +0.12
    0.7%
    240
    +0.12
    0.7%
    381
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    351
    +0.12
    0.09
    13
    +0.12
    0.05
    60
    +0.10
    0.01
    Negative Logits
     ahead
    -1.60
    oltzmann
    -1.42
     yours
    -1.39
    bilt
    -1.38
     affairs
    -1.35
     ours
    -1.35
     minded
    -1.34
     theirs
    -1.32
    entre
    -1.31
     aware
    -1.30
    POSITIVE LOGITS
    atively
    1.59
    arity
    1.53
     rapidly
    1.53
     critically
    1.50
    ij
    1.48
    ¼
    1.46
    ¿
    1.44
    iÄĩ
    1.42
    ga
    1.39
    Ĺ
    1.36
    Activations Density 3.507%

    No Known Activations