© 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. MLP Post
    9. 0-MLP-SM
    11. 25925
    Prev
    Next
    INDEX
    Explanations

    instances of the verb "saw" that indicate perception or observation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__mlp-sm_processed/0-mlp-sm
    Prompts (Dashboard)
    32,768 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.0.hook_mlp_out
    Hook Layer
    0
    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
    376
    +0.20
    1.1%
    115
    +0.16
    0.9%
    148
    +0.15
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    54
    +0.20
    0.04
    235
    +0.16
    0.03
    463
    +0.15
    0.03
    Negative Logits
    ories
    -1.94
    «
    -1.73
    »¿
    -1.44
    MSO
    -1.43
    lor
    -1.42
    robe
    -1.41
    Decided
    -1.40
    clud
    -1.35
    ape
    -1.35
     PBX
    -1.34
    POSITIVE LOGITS
     fit
    1.65
     rid
    1.60
    dust
    1.57
     him
    1.56
     utter
    1.47
    actic
    1.38
     approximate
    1.33
     possible
    1.33
     clear
    1.32
    inal
    1.32
    Activations Density 0.018%

    No Known Activations