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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 29813
    Prev
    Next
    INDEX
    Explanations

    references to seminars or academic events

    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
    18
    +0.16
    0.9%
    494
    +0.14
    0.8%
    218
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    18
    +0.16
    0.03
    386
    +0.14
    0.01
    494
    +0.13
    0.02
    Negative Logits
     consequence
    -1.61
     behalf
    -1.61
    icture
    -1.57
     MERCHANTABILITY
    -1.54
     sender
    -1.46
    breviations
    -1.45
    claimed
    -1.43
    ÅĦ
    -1.42
    coming
    -1.38
    fig
    -1.37
    POSITIVE LOGITS
    Īĺ
    1.85
    ©
    1.71
    inin
    1.62
    Ĥ¬
    1.59
    igroup
    1.58
    ple
    1.56
    ī
    1.55
    ª
    1.55
    esters
    1.54
    keit
    1.53
    Activations Density 0.116%

    No Known Activations