© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    5. GPT2-Small
    7. Residual Similar CE E2E Recon
    9. 6-RES_SCEFR-AJT
    11. 6389
    Prev
    Next
    INDEX
    Explanations

    geometric shapes or spatial relationships

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    neuronpedia/gpt2-small__res_scefr-ajt/6-res_scefr-ajt
    Prompts (Dashboard)
    12,288 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    46,080
    Data Type
    torch.float32
    Hook Point
    blocks.6.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    apollo-research/Skylion007-openwebtext-tokenizer-gpt2
    Hook Point Layer
    6
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    549
    +0.18
    0.7%
    566
    +0.15
    0.6%
    288
    +0.15
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    549
    +0.18
    0.04
    288
    +0.15
    0.06
    1
    +0.15
    0.04
    Negative Logits
    orah
    -0.87
    jab
    -0.85
    offer
    -0.84
    ibur
    -0.83
    inker
    -0.81
    immer
    -0.80
    act
    -0.78
    ository
    -0.77
     hindsight
    -0.76
    hub
    -0.76
    POSITIVE LOGITS
    htaking
    0.82
    Frameworks
    0.81
    ttes
    0.81
    atform
    0.81
    cules
    0.80
     Seym
    0.77
    izabeth
    0.76
    rily
    0.76
    ength
    0.75
    ulkan
    0.75
    Activations Density 0.735%

    No Known Activations