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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 10-RES-JB
    11. 2669
    Prev
    Next
    INDEX
    Explanations

    metaphorical and literal references to hollowness or emptiness

    oai_token-act-pair · gpt-3.5-turbo

    terms related to figurative and literal interpretations, especially focused on "figurative" language and concepts of hollowness

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Downloadha
    -0.74
    board
    -0.74
     guiActiveUnfocused
    -0.73
    boards
    -0.71
    wick
    -0.70
    boarding
    -0.70
    EMENT
    -0.70
    ovic
    -0.70
    keeper
    -0.69
    itism
    -0.67
    POSITIVE LOGITS
    arge
    0.79
    ocity
    0.78
     fulfillment
    0.77
    ative
    0.76
    entric
    0.72
    ©¶æ¥µ
    0.71
    lihood
    0.71
    oyd
    0.70
    aneous
    0.69
    theless
    0.68
    Activations Density 0.080%

    No Known Activations