© 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. 13200
    Prev
    Next
    INDEX
    Explanations

    terms related to cross-referencing in various contexts

    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
    172
    +0.15
    0.8%
    504
    +0.12
    0.6%
    287
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    172
    +0.15
    0.03
    504
    +0.12
    0.02
    399
    +0.11
    0.01
    Negative Logits
    woke
    -1.83
    IRE
    -1.60
    ILITY
    -1.59
    unnumbered
    -1.56
    úblic
    -1.51
    CTOR
    -1.46
    rieve
    -1.46
    ######
    -1.46
    OUNT
    -1.45
    apest
    -1.44
    POSITIVE LOGITS
    Ļª
    2.64
     sectional
    2.16
    ¼
    2.16
    Īĺ
    2.12
    legged
    2.11
    ±
    1.89
    ĭ
    1.87
    roads
    1.86
    ¾
    1.80
    ¹
    1.79
    Activations Density 0.084%

    No Known Activations