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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 12179
    Prev
    Next
    INDEX
    Explanations

    commas or pauses in sentences

    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
    54
    +0.14
    0.8%
    23
    +0.11
    0.6%
    492
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    439
    +0.14
    0.05
    176
    +0.11
    0.06
    2
    +0.11
    0.06
    Negative Logits
    OPLE
    -1.55
    stadt
    -1.41
     unfolded
    -1.40
    lif
    -1.39
    ¿½
    -1.39
    illin
    -1.36
    hood
    -1.35
    ģ
    -1.35
    chair
    -1.34
    ables
    -1.33
    POSITIVE LOGITS
    tig
    2.05
     unlike
    1.69
     luck
    1.66
     contrary
    1.63
    tons
    1.52
    semble
    1.44
     unfortunately
    1.44
    fortunately
    1.41
     nonetheless
    1.41
     since
    1.41
    Activations Density 0.472%

    No Known Activations