INDEX

Explanations

starting phrases like "Core tip:", "Note:", or "*]

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

The neuron flags formatted section labels or headings—i.e. bold or HTML‐wrapped tokens that consist of a label (often a number or phrase) immediately followed by a colon.

oai_token-act-pair · o4-mini

Triggered by @jyhe0408

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_22/width_131k

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

قیمت

-1.24

 řekl

-1.23

ėms

-1.14

灵气

-1.09

ėmis

-1.09

极为

-1.08

دکتر

-1.03

Ɲ

-1.01

ỳnh

-1.01

дной

-1.00

POSITIVE LOGITS

 Gläub

1.39

<strong>

1.36

妡

1.34

the

1.32

纂

1.16

 bekannten

1.15

是想

1.13

 Erfolge

1.13

 ewigen

1.13

 gesetzlichen

1.12

Activations Density 0.004%

starting phrases like "Core tip:", "Note:", or "*]

The neuron flags formatted section labels or headings—i.e. bold or HTML‐wrapped tokens that consist of a label (often a number or phrase) immediately followed by a colon.

No Known Activations

starting phrases like "Core tip:", "Note:", or "*]

The neuron flags formatted section labels or headings—i.e. bold or HTML‐wrapped tokens that consist of a label (often a number or phrase) immediately followed by a colon.

No Known Activations

starting phrases like "**Core tip**:", "**Note:**", or "*]

The neuron flags formatted section labels or headings—i.e. bold or HTML‐wrapped tokens that consist of a label (often a number or phrase) immediately followed by a colon.

No Known Activations

starting phrases like "**Core tip**:", "**Note:**", or "*]

The neuron flags formatted section labels or headings—i.e. bold or HTML‐wrapped tokens that consist of a label (often a number or phrase) immediately followed by a colon.

No Known Activations

starting phrases like "Core tip:", "Note:", or "*]

starting phrases like "Core tip:", "Note:", or "*]