INDEX

Explanations

This neuron fires on the first word (or initial punctuation) at the start of a new sentence or block of text.

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

the word "it" when referring back to something previously mentioned.

oai_token-act-pair · claude-4-5-sonnet Triggered by @jyhe0408

common function words and hedging terms that link clauses or express uncertainty (pronouns, prepositions, conjunctions, and modifiers).

oai_token-act-pair · gpt-5 Triggered by @jyhe0408

New Auto-Interp

Configuration

google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

可以选择

0.80

series

0.74

 Ді

0.74

三种

0.71

 选择

0.71

 সাত

0.70

 类型

0.70

两

0.70

它

0.69

荤

0.69

POSITIVE LOGITS

ی

0.92

ה

0.86

 dica

0.79

 Lorena

0.79

عهد

0.76

 haft

0.75

 dement

0.74

bei

0.74

 klin

0.74

 หน่วย

0.73

Activations Density 0.000%

This neuron fires on the first word (or initial punctuation) at the start of a new sentence or block of text.

the word "it" when referring back to something previously mentioned.

common function words and hedging terms that link clauses or express uncertainty (pronouns, prepositions, conjunctions, and modifiers).

No Comments

No Known Activations

This neuron fires on the first word (or initial punctuation) at the start of a new sentence or block of text.

the word "it" when referring back to something previously mentioned.

common function words and hedging terms that link clauses or express uncertainty (pronouns, prepositions, conjunctions, and modifiers).

No Comments

No Known Activations