INDEX

Explanations

numbers

np_max-act · gemini-2.0-flash

The neuron strongly activates on numeric tokens—especially decimal numbers (e.g. hazard ratios, percentages, counts)—highlighting statistical values in the text.

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

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

 Arts

-0.07

罗斯

-0.06

road

-0.06

 Mars

-0.06

.Count

-0.06

.SetInt

-0.06

LD

-0.06

.Line

-0.06

Road

-0.06

 Language

-0.06

POSITIVE LOGITS

 신입

0.07

"strings

0.07

implements

0.06

	before

0.06

ParseException

0.06

Tek

0.06

 الحل

0.06

uffy

0.06

其他

0.06

 згідно

0.06

Activations Density 0.075%

numbers

The neuron strongly activates on numeric tokens—especially decimal numbers (e.g. hazard ratios, percentages, counts)—highlighting statistical values in the text.

No Comments

No Known Activations