© 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. 9673
    Prev
    Next
    INDEX
    Explanations

    technical terms and references related to computer programming or software documentation

    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
    156
    +0.16
    0.9%
    181
    +0.13
    0.7%
    159
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    181
    +0.16
    0.13
    450
    +0.13
    0.10
    433
    +0.13
    0.05
    Negative Logits
     cooker
    -1.62
    /#
    -1.41
    caster
    -1.36
     technique
    -1.33
     repertoire
    -1.31
    /-/
    -1.31
    Selector
    -1.28
     downstairs
    -1.27
     amplifier
    -1.27
    Accessor
    -1.27
    POSITIVE LOGITS
    **^
    1.59
    ights
    1.55
     licenses
    1.50
    rapeutic
    1.41
    ^**
    1.41
    ĻĤ
    1.39
    ety
    1.38
    rotic
    1.36
     lives
    1.34
     lic
    1.34
    Activations Density 4.142%

    No Known Activations