INDEX

Explanations

passes and Triple

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

The neuron is triggered by sequences of repeated punctuation characters (e.g. “===”, “---”, “///” or long runs of “—” or “–”) typically used as section/divider lines in code comments or headers.

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

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_34/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

Ado

-0.76

 भेज

-0.73

 سين

-0.71

Mes

-0.70

zeka

-0.67

 Yash

-0.66

 Slaves

-0.66

comber

-0.65

गा

-0.64

PURE

-0.64

POSITIVE LOGITS

 pass

1.12

 passes

1.06

pass

1.05

 Pass

0.98

 Passes

0.94

Pass

0.89

Passes

0.87

 Triple

0.85

passes

0.84

utique

0.79

Activations Density 0.126%

passes and Triple

The neuron is triggered by sequences of repeated punctuation characters (e.g. “===”, “---”, “///” or long runs of “—” or “–”) typically used as section/divider lines in code comments or headers.

No Comments

No Known Activations

passes and Triple

The neuron is triggered by sequences of repeated punctuation characters (e.g. “===”, “---”, “///” or long runs of “—” or “–”) typically used as section/divider lines in code comments or headers.

No Comments

No Known Activations

passes and Triple

The neuron is triggered by sequences of repeated punctuation characters (e.g. “===”, “---”, “///” or long runs of “—” or “*–*”) typically used as section/divider lines in code comments or headers.

No Comments

No Known Activations

passes and Triple

The neuron is triggered by sequences of repeated punctuation characters (e.g. “===”, “---”, “///” or long runs of “—” or “*–*”) typically used as section/divider lines in code comments or headers.

No Comments

No Known Activations

The neuron is triggered by sequences of repeated punctuation characters (e.g. “===”, “---”, “///” or long runs of “—” or “–”) typically used as section/divider lines in code comments or headers.

The neuron is triggered by sequences of repeated punctuation characters (e.g. “===”, “---”, “///” or long runs of “—” or “–”) typically used as section/divider lines in code comments or headers.