© 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. 12942
    Prev
    Next
    INDEX
    Explanations

    mentions of Donald Trump and related figures, indicating a focus on political references

    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
    352
    +0.16
    0.9%
    410
    +0.15
    0.9%
    156
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    352
    +0.16
    0.03
    410
    +0.15
    0.03
    251
    +0.12
    0.02
    Negative Logits
    Ļª
    -1.94
    ĥ½
    -1.76
    idence
    -1.64
    Ĺ
    -1.55
    rid
    -1.55
    EOF
    -1.54
    ¿½
    -1.54
    rapeut
    -1.52
    olate
    -1.51
    ĥ
    -1.51
    POSITIVE LOGITS
    sson
    1.89
    ovich
    1.88
    stein
    1.81
    son
    1.68
    igan
    1.66
     Jr
    1.63
    ford
    1.55
     Trump
    1.54
    pson
    1.54
    hardt
    1.53
    Activations Density 0.126%

    No Known Activations