© 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. 30134
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "new"

    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
    477
    +0.13
    0.7%
    58
    +0.13
    0.7%
    313
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    66
    +0.13
    0.05
    58
    +0.13
    0.04
    477
    +0.12
    0.05
    Negative Logits
    onitrile
    -1.56
    FFIR
    -1.42
     bacter
    -1.40
     overflow
    -1.36
    ureau
    -1.35
    UGH
    -1.35
    abol
    -1.32
    izens
    -1.31
     bacteri
    -1.30
     climax
    -1.29
    POSITIVE LOGITS
    ±
    3.35
    ¡
    3.28
    ¨
    3.25
    ¦
    3.23
    ĻĤ
    3.18
    ¸
    3.15
    ³
    3.14
    Ģ
    3.08
    ¹
    3.06
    ĭ
    3.01
    Activations Density 0.564%

    No Known Activations