INDEX

Explanations

their

np_max-act · gemini-2.0-flash

The neuron fires on comparative‐type specification words—e.g. “different,” “their,” “interactions,” “based”—essentially marking phrases that explain separation criteria.

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

]["

-0.07

Та

-0.06

	cell

-0.06

.on

-0.06

 broker

-0.06

Pok

-0.06

 nuisance

-0.06

ague

-0.06

}"↵

-0.06

ERCHANT

-0.06

POSITIVE LOGITS

 Bukkit

0.06

 disillusion

0.06

 Josef

0.06

 رض

0.06

�

0.06

dg

0.06

(priority

0.06

(sw

0.06

 sausage

0.06

 دفاع

0.06

Activations Density 0.017%

their

The neuron fires on comparative‐type specification words—e.g. “different,” “their,” “interactions,” “based”—essentially marking phrases that explain separation criteria.

No Comments

No Known Activations