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. 45145
    Prev
    Next
    INDEX
    Explanations

    terms related to trust, distrust, suspicion, and confidence in various contexts

    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
    +0.08
    0.7%
    288
    +0.07
    0.6%
    255
    +0.07
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.08
    0.23
    641
    +0.07
    0.12
    225
    +0.07
    0.10
    Negative Logits
    cise
    -0.34
     Violet
    -0.33
     Kepler
    -0.31
    iolet
    -0.31
    usha
    -0.30
     Genius
    -0.29
    uscript
    -0.28
     Parenthood
    -0.28
    detail
    -0.28
     revolutions
    -0.27
    POSITIVE LOGITS
    worthiness
    0.77
    fulness
    0.54
    lessness
    0.50
    worthy
    0.48
    ful
    0.45
    ability
    0.45
    lessly
    0.43
     trust
    0.41
    phal
    0.41
    lust
    0.40
    Activations Density 5.757%

    No Known Activations