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. 44777
    Prev
    Next
    INDEX
    Explanations

    personal experiences and reflections

    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
    +0.27
    1.9%
    481
    +0.10
    0.7%
    393
    +0.09
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    566
    +0.27
    0.18
    551
    +0.10
    0.16
    149
    +0.09
    0.10
    Negative Logits
    berra
    -0.42
     Peak
    -0.39
    advertisement
    -0.38
    acher
    -0.38
    ultan
    -0.37
    eeds
    -0.36
    iversary
    -0.36
     Sung
    -0.35
     interfering
    -0.34
    wake
    -0.34
    POSITIVE LOGITS
    'll
    0.64
     hereby
    0.60
    'd
    0.51
    've
    0.49
     reasoned
    0.49
     wondered
    0.47
     sugg
    0.44
     must
    0.44
    bra
    0.44
     nevertheless
    0.42
    Activations Density 12.006%

    No Known Activations