INDEX

Explanations

words starting with 'wa'

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

The neuron is highly sensitive to letter sequences beginning with “wa” (as in “war,” “wan,” “wai,” etc.)—i.e. it spots any token containing the substring “wa.”

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

轳

-3.19

粜

-3.06

-2.92

-2.70

-2.67

刄

-2.63

 hordes

-2.61

 ไม่

-2.53

簠

-2.53

很少

-2.50

POSITIVE LOGITS

！”

2.45

 古风

2.42

！！”

2.31

 хотела

2.31

ቼ

2.25

数量

2.13

！！！”

2.11

！（

2.11

 mariana

2.11

 unrelenting

2.09

Activations Density 0.019%

words starting with 'wa'

The neuron is highly sensitive to letter sequences beginning with “wa” (as in “war,” “wan,” “wai,” etc.)—i.e. it spots any token containing the substring “wa.”

No Comments

No Known Activations

words starting with 'wa'

The neuron is highly sensitive to letter sequences beginning with “wa” (as in “war,” “wan,” “wai,” etc.)—i.e. it spots any token containing the substring “wa.”

No Comments

No Known Activations

words starting with 'wa'

The neuron is highly sensitive to letter sequences beginning with “w­a” (as in “war,” “wan,” “wai,” etc.)—i.e. it spots any token containing the substring “wa.”

No Comments

No Known Activations

words starting with 'wa'

The neuron is highly sensitive to letter sequences beginning with “w­a” (as in “war,” “wan,” “wai,” etc.)—i.e. it spots any token containing the substring “wa.”

No Comments

No Known Activations

The neuron is highly sensitive to letter sequences beginning with “wa” (as in “war,” “wan,” “wai,” etc.)—i.e. it spots any token containing the substring “wa.”

The neuron is highly sensitive to letter sequences beginning with “wa” (as in “war,” “wan,” “wai,” etc.)—i.e. it spots any token containing the substring “wa.”