© 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. 17960
    Prev
    Next
    INDEX
    Explanations

    phrases or sentences that are worded in a way that presents a particular perspective or viewpoint

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "framed" in various contexts

    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
     pickups
    -0.96
    ynes
    -0.86
     Naj
    -0.82
    Berry
    -0.76
    ragon
    -0.75
    ois
    -0.74
    aires
    -0.72
    ancy
    -0.71
    ules
    -0.71
    uper
    -0.71
    POSITIVE LOGITS
     framed
    3.07
     framing
    2.41
     Fram
    1.62
    fram
    1.58
     fram
    1.51
     Rig
    1.06
    soType
    1.04
     Rim
    0.91
     polarization
    0.90
    border
    0.89
    Activations Density 0.046%

    No Known Activations