© 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. 0-RES-JB
    11. 11674
    Prev
    Next
    INDEX
    Explanations

    the word "Murd" or "Murd" within a larger context

    oai_token-act-pair · gpt-3.5-turbo

    references to murder or related themes

    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
    cess
    -0.82
    gression
    -0.74
    bos
    -0.71
    ubs
    -0.68
     Tier
    -0.68
     brakes
    -0.67
    etics
    -0.65
     timing
    -0.65
    CG
    -0.65
    atre
    -0.63
    POSITIVE LOGITS
     Murd
    4.12
    mur
    1.32
    Mur
    1.29
     Murder
    1.27
     Massacre
    1.24
     murd
    1.24
     murders
    1.19
     Lect
    1.15
     landslide
    1.06
     Daredevil
    1.06
    Activations Density 0.049%

    No Known Activations