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. 13297
    Prev
    Next
    INDEX
    Explanations

    mentions of film-related terms, specifically cinematography and camera work

    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
    167
    +0.08
    0.5%
    38
    +0.07
    0.5%
    555
    +0.07
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    555
    +0.08
    0.07
    167
    +0.07
    -0.04
    763
    +0.07
    0.07
    Negative Logits
     Curiosity
    -0.47
     hydra
    -0.45
     Predict
    -0.45
     tolerant
    -0.44
     Panther
    -0.44
     Civic
    -0.43
     futile
    -0.43
     suscept
    -0.42
     hatch
    -0.42
     embold
    -0.42
    POSITIVE LOGITS
    atography
    0.84
    atur
    0.70
    ographer
    0.70
    amen
    0.70
    ographed
    0.70
    ago
    0.68
    icians
    0.67
    ographers
    0.65
    artisan
    0.65
    odynamic
    0.64
    Activations Density 9.324%

    No Known Activations