© 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. 28539
    Prev
    Next
    INDEX
    Explanations

    various forms of the word "from" in relation to comparisons and origins

    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
    219
    +0.12
    0.7%
    485
    +0.11
    0.6%
    155
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    300
    +0.12
    0.09
    409
    +0.11
    0.11
    176
    +0.11
    0.10
    Negative Logits
     interim
    -1.60
    ium
    -1.54
    Lett
    -1.47
    *(
    -1.42
    ered
    -1.41
    cember
    -1.36
    Ins
    -1.34
     involved
    -1.33
    staff
    -1.32
    esc
    -1.32
    POSITIVE LOGITS
    ĥ½
    3.50
    ŀ
    3.48
    ĨĴ
    3.38
    IJ
    3.37
    č↵
    3.32
    ↵
    3.32
              ↵
    3.32
    ↵
    3.32
    ↵
    3.32
    3.32
    Activations Density 1.557%

    No Known Activations