© 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. 13664
    Prev
    Next
    INDEX
    Explanations

    phrases related to describing capabilities or achievements in a variety of 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
    24
    +0.14
    0.8%
    78
    +0.13
    0.8%
    271
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    27
    +0.14
    0.08
    24
    +0.13
    0.09
    271
    +0.13
    -0.05
    Negative Logits
     vill
    -1.50
     Administration
    -1.42
     Judiciary
    -1.40
     chambers
    -1.40
    ubern
    -1.40
    chaft
    -1.39
    rå
    -1.38
     Organization
    -1.36
     etc
    -1.35
     tors
    -1.35
    POSITIVE LOGITS
    osph
    1.53
    Ī
    1.48
     closely
    1.44
    née
    1.44
     awhile
    1.43
    aporation
    1.43
     immediately
    1.42
    оваÑĤÑĮ
    1.40
    »
    1.40
     cautiously
    1.36
    Activations Density 2.966%

    No Known Activations