© 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
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 2-RES-SM
    11. 11107
    Prev
    Next
    INDEX
    Explanations

    references to types of plants and food, specifically beans

    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
    +0.20
    1.1%
    43
    +0.17
    1.0%
    38
    +0.15
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    410
    +0.20
    0.05
    43
    +0.17
    0.02
    38
    +0.15
    0.02
    Negative Logits
    aneous
    -1.57
    ppen
    -1.50
    ff
    -1.45
    aneously
    -1.44
    ../
    -1.43
    cludes
    -1.37
    icable
    -1.36
    aturally
    -1.35
     equip
    -1.34
    ../../
    -1.34
    POSITIVE LOGITS
    roots
    1.84
    elijk
    1.80
    owski
    1.74
    stalk
    1.67
    holder
    1.66
    dom
    1.60
    hardt
    1.57
    wood
    1.57
    leys
    1.55
    root
    1.54
    Activations Density 0.417%

    No Known Activations