© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 18799
    Prev
    Next
    INDEX
    Explanations

    words related to change or transformation

    oai_token-act-pair · gpt-3.5-turbo

    forms of the word "solve" or related terms that imply resolution or problem-solving

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     mattress
    -0.66
     nap
    -0.65
     shower
    -0.64
     Socialism
    -0.62
     bunk
    -0.61
     warnings
    -0.59
     Warn
    -0.58
     stuffing
    -0.57
     reading
    -0.57
     conspicuous
    -0.56
    POSITIVE LOGITS
    olved
    4.08
    olve
    3.96
    olving
    3.27
    olves
    3.22
    olution
    2.67
    olutions
    2.28
    olver
    1.92
    olute
    1.21
    atisf
    1.11
    ourced
    1.08
    Activations Density 0.016%

    No Known Activations