© 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. 30620
    Prev
    Next
    INDEX
    Explanations

    mathematical expressions and equations

    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
    369
    +0.17
    1.0%
    436
    +0.11
    0.6%
    126
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    81
    +0.17
    0.02
    369
    +0.11
    0.01
    436
    +0.11
    0.02
    Negative Logits
     whom
    -1.65
     paste
    -1.51
    "?
    -1.46
     hypertension
    -1.45
    ?"
    -1.45
    ?_
    -1.43
    ↵
    -1.37
     stuff
    -1.36
    Ĥ¬
    -1.36
     burden
    -1.35
    POSITIVE LOGITS
    Docket
    1.65
    ête
    1.49
     Appeal
    1.48
    ymphony
    1.46
    jem
    1.43
    EEK
    1.43
    Appeal
    1.38
    olini
    1.33
     Romanian
    1.31
    Italian
    1.31
    Activations Density 0.045%

    No Known Activations