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

    Neuronpedia

    APICircuit 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 L0 E2E Recon
    5. 2-RES_SLEFR-AJT
    6. 5450
    Prev
    Next
    INDEX
    Explanations

    adjectives and descriptors related to quality and appearance

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    neuronpedia/gpt2-small__res_slefr-ajt/2-res_slefr-ajt
    Prompts (Dashboard)
    12,288 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    46,080
    Data Type
    torch.float32
    Hook Point
    blocks.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    apollo-research/Skylion007-openwebtext-tokenizer-gpt2
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    257
    +0.06
    0.5%
    158
    +0.06
    0.5%
    11
    +0.06
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    555
    +0.06
    0.01
    326
    +0.06
    0.01
    195
    +0.06
    0.00
    Negative Logits
    orio
    -0.52
    ĪĴ
    -0.49
    ooter
    -0.47
    pmwiki
    -0.45
    ©¶æ
    -0.44
    initions
    -0.43
    olars
    -0.42
    heading
    -0.40
    BACK
    -0.40
    occupied
    -0.40
    POSITIVE LOGITS
     Atmosp
    0.44
     Seym
    0.44
     Brewer
    0.38
     Hawk
    0.38
     Fresh
    0.37
    tis
    0.37
     Prosper
    0.36
     Mell
    0.36
     Norn
    0.36
    helle
    0.36
    Activations Density 0.007%

    No Known Activations