Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 29272
    Prev
    Next
    INDEX
    Explanations

    code elements related to variable assignments and comparisons

    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
    478
    +0.12
    0.7%
    115
    +0.12
    0.7%
    369
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    497
    +0.12
    0.03
    133
    +0.12
    0.05
    345
    +0.12
    0.06
    Negative Logits
    ieri
    -1.84
    izer
    -1.75
    eries
    -1.55
    iously
    -1.53
    ously
    -1.46
    ientos
    -1.44
     absor
    -1.44
    Listener
    -1.43
     [(
    -1.43
    ector
    -1.43
    POSITIVE LOGITS
    ĨĴ
    2.60
    ī
    2.54
    ³
    2.38
    Ĺ
    2.34
    ¬
    2.32
    ¿
    2.29
    ¿½
    2.28
    ¦
    2.28
    ¤
    2.24
    Īĺ
    2.22
    Activations Density 1.120%

    No Known Activations