INDEX

Explanations

code structure analysis

np_acts-logits-general · gemini-2.5-flash-lite

closing parentheses followed by opening curly braces in code, typically marking the start of function or method bodies.

oai_token-act-pair · claude-4-5-sonnet Triggered by @bfgcn7frf5

The neuron fires strongly on tokens that introduce or open a code block or function (e.g. “=>” in JavaScript arrow functions, “extern "C"” in Rust/C interop, and “do” in Ruby/RSpec).

oai_token-act-pair · o4-mini Triggered by @jyhe0408

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_10/width_131k

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

のですね

-1.59

-1.41

ようですが

-1.41

-1.34

できるだけ

-1.29

際には

-1.21

かったです

-1.20

 blanches

-1.20

 quarante

-1.17

のかもしれません

-1.17

POSITIVE LOGITS

and

2.11

 funktioniert

1.66

 wichtig

1.47

忽然

1.45

becca

1.44

 doua

1.42

\&

1.41

\&

1.41

 ਹੈ

1.39

 funcione

1.36

Activations Density 0.054%

code structure analysis

closing parentheses followed by opening curly braces in code, typically marking the start of function or method bodies.

The neuron fires strongly on tokens that introduce or open a code block or function (e.g. “=>” in JavaScript arrow functions, “extern "C"” in Rust/C interop, and “do” in Ruby/RSpec).

No Comments

No Known Activations

code structure analysis

closing parentheses followed by opening curly braces in code, typically marking the start of function or method bodies.

The neuron fires strongly on tokens that introduce or open a code block or function (e.g. “=>” in JavaScript arrow functions, “extern "C"” in Rust/C interop, and “do” in Ruby/RSpec).

No Comments

No Known Activations