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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 13224
    Prev
    Next
    INDEX
    Explanations

    mentions of legal procedures and government actions

    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
    36
    +0.26
    1.1%
    393
    +0.22
    0.9%
    447
    +0.17
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    36
    +0.26
    0.35
    362
    +0.22
    0.20
    393
    +0.17
    0.25
    Negative Logits
    evidence
    -0.91
     heirs
    -0.79
    iments
    -0.73
    encies
    -0.73
    Enjoy
    -0.72
     proceedings
    -0.71
    Events
    -0.71
    ographs
    -0.71
     honors
    -0.71
     Amend
    -0.70
    POSITIVE LOGITS
     bunch
    1.14
     whopping
    1.12
     handful
    1.08
     lot
    1.08
     certain
    1.05
     few
    1.04
     variety
    1.03
     multitude
    1.03
     similar
    1.02
     single
    1.02
    Activations Density 7.350%

    No Known Activations