© 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. 12921
    Prev
    Next
    INDEX
    Explanations

    value and usability in products and processes

    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
    263
    +0.23
    1.4%
    56
    +0.17
    1.0%
    153
    +0.15
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    263
    +0.23
    0.16
    56
    +0.17
    -0.03
    141
    +0.15
    0.08
    Negative Logits
     Members
    -1.60
     ADP
    -1.52
     Member
    -1.47
     hostage
    -1.47
     Interests
    -1.45
     IRA
    -1.42
     Tud
    -1.36
     Hod
    -1.35
     Register
    -1.35
     representing
    -1.32
    POSITIVE LOGITS
    ureus
    1.79
    semb
    1.76
    sure
    1.59
    fefe
    1.56
    idenote
    1.53
     sake
    1.48
    bersome
    1.45
    Argued
    1.44
    skip
    1.43
     faults
    1.41
    Activations Density 3.798%

    No Known Activations