INDEX

Explanations

first person

np_max-act · gemini-2.0-flash

The neuron strongly activates on first‐person “I” statements expressing user intentions (e.g. “I want to…,” “I’ll give you…,” “I am…”).

oai_token-act-pair · o4-mini Triggered by @xinyanhu8

first-person self-referential phrasing (user statements/questions using self-focused auxiliaries), especially near the start of the utterance.

oai_token-act-pair · gpt-5 Triggered by @vetterc0

New Auto-Interp

Configuration

andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_11/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.11.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ByVersion

-0.07

xv

-0.06

Tor

-0.06

.section

-0.06

Le

-0.06

tti

-0.06

 simplex

-0.06

encv

-0.06

_SHADER

-0.06

Coroutine

-0.06

POSITIVE LOGITS

\E

0.07

(s

0.06

склад

0.06

 payer

0.06

verages

0.06

讀

0.06

-trans

0.06

 CONTRIBUT

0.06

 wrists

0.06

 amendments

0.06

Activations Density 0.012%

first person

The neuron strongly activates on first‐person “I” statements expressing user intentions (e.g. “I want to…,” “I’ll give you…,” “I am…”).

first-person self-referential phrasing (user statements/questions using self-focused auxiliaries), especially near the start of the utterance.

No Comments

No Known Activations

first person

The neuron strongly activates on first‐person “I” statements expressing user intentions (e.g. “I want to…,” “I’ll give you…,” “I am…”).

first-person self-referential phrasing (user statements/questions using self-focused auxiliaries), especially near the start of the utterance.

No Comments

No Known Activations