© 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
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 8-RES-JB
    11. 7876
    Prev
    Next
    INDEX
    Explanations

    phrases related to placing entities in certain positions or situations

    oai_token-act-pair · gpt-3.5-turbo

    phrases that indicate position or status in relation to various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     uttered
    -0.66
     motions
    -0.64
     refuses
    -0.63
     forbids
    -0.60
     happens
    -0.60
     manifests
    -0.60
     (%
    -0.59
    leys
    -0.58
     consists
    -0.57
     wisely
    -0.57
    POSITIVE LOGITS
     jeopardy
    0.94
    ãĥ¯ãĥ³
    0.91
    pmwiki
    0.79
    ãĤ´ãĥ³
    0.74
     peril
    0.73
     unwelcome
    0.71
    ãĤ§
    0.71
     unch
    0.68
    rongh
    0.68
     exha
    0.66
    Activations Density 0.186%

    No Known Activations