INDEX

Explanations

chopped

np_max-act · gemini-2.0-flash

The neuron activates on past‐participle cooking prep words (e.g. “chopped,” “diced,” “grated”) indicating ingredient preparation steps.

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

 الأخرى

-0.06

рует

-0.06

_SITE

-0.06

REL

-0.06

 submits

-0.06

.request

-0.06

 Mormons

-0.06

สำเร

-0.06

(NS

-0.06

redict

-0.06

POSITIVE LOGITS

 chopped

0.09

裁

0.07

 chopping

0.07

.Department

0.07

 وف

0.07

 pomoc

0.07

nin

0.07

 چاپ

0.07

chap

0.07

Activations Density 0.004%

chopped

The neuron activates on past‐participle cooking prep words (e.g. “chopped,” “diced,” “grated”) indicating ingredient preparation steps.

No Comments

No Known Activations