INDEX

Explanations

biological and technical terms

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

nouns followed by verbs in technical or scientific contexts.

oai_token-act-pair · claude-3-7-sonnet-20250219

Triggered by @neilrathi

The neuron fires strongly on numerals, numeric sequences, or tokens containing digits (e.g. counts, exam numbers, measurements).

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

Negative Logits

 when

-1.49

if

-1.46

nachron

-1.33

etna

-1.30

phrodite

-1.30

 therefore

-1.25

 that

-1.22

 mahdol

-1.20

maakt

-1.20

madeus

-1.20

POSITIVE LOGITS

 nawet

1.41

it

1.36

无论是

1.29

is

1.23

[],

1.21

we

1.20

these

1.17

 unirse

1.16

things

1.16

this

1.13

Activations Density 0.255%

biological and technical terms

nouns followed by verbs in technical or scientific contexts.

The neuron fires strongly on numerals, numeric sequences, or tokens containing digits (e.g. counts, exam numbers, measurements).

No Known Activations

biological and technical terms

nouns followed by verbs in technical or scientific contexts.

The neuron fires strongly on numerals, numeric sequences, or tokens containing digits (e.g. counts, exam numbers, measurements).

No Known Activations