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. 34126
    Prev
    Next
    INDEX
    Explanations

    descriptions of mythical or fantastical creatures

    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
    322
    +0.13
    0.5%
    717
    +0.12
    0.5%
    31
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    701
    +0.13
    0.05
    406
    +0.12
    0.07
    717
    +0.12
    0.05
    Negative Logits
    advertisement
    -0.77
     datas
    -0.76
    writers
    -0.75
     merciless
    -0.74
    sheets
    -0.74
     leash
    -0.73
    lining
    -0.70
     thirds
    -0.69
     Rhod
    -0.67
    glers
    -0.67
    POSITIVE LOGITS
    lehem
    1.20
    ixture
    1.01
    ibo
    0.98
    achusetts
    0.98
    ments
    0.95
    assador
    0.93
    akeru
    0.92
    ixtures
    0.91
    pas
    0.91
    estic
    0.91
    Activations Density 6.176%

    No Known Activations