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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsBlogSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    3. GPT2-Small
    4. Residual Similar CE Local
    5. 6-RES_SCL-AJT
    6. 40134
    Prev
    Next
    INDEX
    Explanations

    instances of positions being appointed or replaced

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    neuronpedia/gpt2-small__res_scl-ajt/6-res_scl-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
    447
    +0.20
    0.9%
    101
    +0.14
    0.6%
    339
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.20
    0.32
    385
    +0.14
    0.20
    362
    +0.13
    0.16
    Negative Logits
    patch
    -0.85
    Ïģ
    -0.79
    λ
    -0.76
    iday
    -0.74
    pack
    -0.74
    pling
    -0.74
    psc
    -0.70
    fal
    -0.70
    olver
    -0.68
    iland
    -0.67
    POSITIVE LOGITS
     appointments
    1.02
     appoint
    0.92
    ees
    0.90
     appointment
    0.89
     appointed
    0.86
    ointed
    0.85
    eering
    0.84
     ineligible
    0.84
    ancies
    0.82
    lies
    0.80
    Activations Density 14.615%

    No Known Activations