© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 4-RES-SM
    6. 11765
    Prev
    Next
    INDEX
    Explanations

    references to dropdown menus or selection options in a user interface context

    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
    217
    +0.12
    0.7%
    307
    +0.11
    0.6%
    168
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    164
    +0.12
    0.03
    120
    +0.11
    0.03
    437
    +0.10
    0.03
    Negative Logits
    ities
    -2.10
     Confed
    -1.83
    oral
    -1.73
    ity
    -1.70
    itarian
    -1.59
    ariable
    -1.48
    eren
    -1.48
    isd
    -1.46
    ITY
    -1.45
    acular
    -1.45
    POSITIVE LOGITS
    out
    2.36
    kick
    2.32
    down
    2.20
    outs
    2.16
     downs
    2.04
    zone
    2.01
    Down
    1.99
    box
    1.95
    ®
    1.95
     down
    1.86
    Activations Density 0.250%

    No Known Activations