INDEX

Explanations

markers, pointers, or delimiters

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

The neuron responds to the “<-” (and similar left‐arrow) tokens used as in‐line pointers or markers in code comments and diffs.

oai_token-act-pair · o4-mini

Triggered by @jyhe0408

inline arrow/chevron markers used as annotations in code or console snippets (e.g., left/right arrows and HTML-style comment delimiters)

oai_token-act-pair · gpt-5

Triggered by @jyhe0408

arrows or pointers in code comments indicating references or annotations.

oai_token-act-pair · claude-4-5-sonnet

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

Negative Logits

-1.51

热的

-1.48

襯

-1.45

灕

-1.44

夢見

-1.44

耸

-1.41

见的

-1.39

舒適

-1.38

ܜ

-1.38

という感じ

-1.36

POSITIVE LOGITS

2.06

.'

1.98

</i>

1.92

1.65

 assist

1.50

Ꮗ

1.49

 właśnie

1.47

tividade

1.47

.,

1.46

'),

1.45

Activations Density 0.024%

markers, pointers, or delimiters

The neuron responds to the “<-” (and similar left‐arrow) tokens used as in‐line pointers or markers in code comments and diffs.

inline arrow/chevron markers used as annotations in code or console snippets (e.g., left/right arrows and HTML-style comment delimiters)

arrows or pointers in code comments indicating references or annotations.

No Known Activations

markers, pointers, or delimiters

The neuron responds to the “<-” (and similar left‐arrow) tokens used as in‐line pointers or markers in code comments and diffs.

inline arrow/chevron markers used as annotations in code or console snippets (e.g., left/right arrows and HTML-style comment delimiters)

arrows or pointers in code comments indicating references or annotations.

No Known Activations