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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 3-RES-JB
    11. 16064
    Prev
    Next
    INDEX
    Explanations

    excerpts relating to caution or being wary

    oai_token-act-pair · gpt-3.5-turbo

    repeated mentions of the word "careful."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    NZ
    -0.81
    SN
    -0.78
    Stars
    -0.77
    CVE
    -0.74
    upon
    -0.73
    flat
    -0.73
    è¦ļéĨĴ
    -0.73
    Wars
    -0.72
    NT
    -0.72
    artifacts
    -0.72
    POSITIVE LOGITS
     calibr
    0.91
     scrutiny
    0.88
    tarian
    0.84
    taker
    0.81
     heed
    0.80
     careful
    0.78
     stewards
    0.76
    ness
    0.73
     empir
    0.72
     deliber
    0.71
    Activations Density 0.016%

    No Known Activations