© 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
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 28766
    Prev
    Next
    INDEX
    Explanations

    concepts related to military service and the sacrifices made by service members for their communities and loved ones

    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
    83
    +0.16
    0.9%
    198
    +0.16
    0.9%
    186
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    14
    +0.16
    0.17
    176
    +0.16
    0.16
    83
    +0.12
    0.06
    Negative Logits
    >;
    -1.50
    ème
    -1.40
    </
    -1.39
    eks
    -1.37
    enden
    -1.34
     concer
    -1.32
    :</
    -1.32
     Chan
    -1.31
    docs
    -1.31
     illustr
    -1.31
    POSITIVE LOGITS
     defense
    1.73
     recovery
    1.53
    hift
    1.49
     preservation
    1.41
    hee
    1.40
    ullah
    1.38
    º
    1.37
     safe
    1.37
    asing
    1.36
     resilient
    1.36
    Activations Density 7.275%

    No Known Activations