INDEX

Explanations

true

np_max-act · gemini-2.0-flash

The neuron flags occurrences of the word “true” in contexts describing that allegations or facts are “taken as true” under legal pleading or summary‐judgment standards.

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

징

-0.07

etiyle

-0.07

	hash

-0.07

.Schedule

-0.06

plr

-0.06

ennai

-0.06

~/

-0.06

hapus

-0.06

 Hiç

-0.06

uben

-0.06

POSITIVE LOGITS

егда

0.07

��

0.07

akedown

0.07

ção

0.06

 Trần

0.06

 Reporting

0.06

extends

0.06

 POSITION

0.06

GT

0.06

 copyright

0.06

Activations Density 0.000%

true

The neuron flags occurrences of the word “true” in contexts describing that allegations or facts are “taken as true” under legal pleading or summary‐judgment standards.

No Comments

No Known Activations