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. 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. 33603
    Prev
    Next
    INDEX
    Explanations

    mentions of different volumes or collections

    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
    +0.22
    1.1%
    271
    +0.13
    0.7%
    288
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    1
    +0.22
    0.09
    375
    +0.13
    0.03
    271
    +0.12
    0.03
    Negative Logits
    unes
    -0.70
     Pearce
    -0.69
    yg
    -0.66
     Turing
    -0.66
    athing
    -0.65
    notations
    -0.65
    jay
    -0.62
    Ķ
    -0.62
    mond
    -0.62
    orius
    -0.62
    POSITIVE LOGITS
     volumes
    0.86
    umes
    0.84
     volume
    0.79
    catentry
    0.78
    Volume
    0.75
     discounts
    0.73
     knob
    0.68
    volume
    0.68
    Increases
    0.67
    é¾
    0.67
    Activations Density 9.037%

    No Known Activations