INDEX

Explanations

biological membranes

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

African language names and place names.

oai_token-act-pair · claude-3-7-sonnet-20250219 Triggered by @neilrathi

The neuron responds strongly to tokens that begin with a capital letter—i.e. it flags uppercase or title‐cased tokens (often names, headings, or copyright/license markers).

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

in

-0.96

髑

-0.93

を

-0.93

義務

-0.89

憩

-0.88

轍

-0.86

姪

-0.84

 मु

-0.84

obatan

-0.84

詣

-0.83

POSITIVE LOGITS

つづく

1.07

んですよね

1.06

んですね

0.98

ETIES

0.93

 bose

0.92

 dizaines

0.92

 milliers

0.92

たびに

0.91

ことがあります

0.90

 riguarda

0.89

Activations Density 0.001%

biological membranes

African language names and place names.

The neuron responds strongly to tokens that begin with a capital letter—i.e. it flags uppercase or title‐cased tokens (often names, headings, or copyright/license markers).

No Comments

No Known Activations

biological membranes

African language names and place names.

The neuron responds strongly to tokens that begin with a capital letter—i.e. it flags uppercase or title‐cased tokens (often names, headings, or copyright/license markers).

No Comments

No Known Activations