© 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. 4-RES-JB
    11. 10192
    Prev
    Next
    INDEX
    Explanations

    instances of the word "put" followed by a number that describes the action related to placing or arranging something

    oai_token-act-pair · gpt-3.5-turbo

    the phrase "put" in various contexts

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

    No Comments

    Negative Logits
    externalActionCode
    -0.96
     Hear
    -0.65
    riott
    -0.64
    reader
    -0.64
    ording
    -0.64
    DAQ
    -0.64
    lus
    -0.61
    SPONSORED
    -0.61
    bie
    -0.61
     --------------------------------
    -0.59
    POSITIVE LOGITS
    ongh
    0.89
     forth
    0.85
    tin
    0.83
     together
    0.82
    tering
    0.78
    rid
    0.76
    otom
    0.74
     toget
    0.74
    tered
    0.73
    ronics
    0.73
    Activations Density 0.041%

    No Known Activations