INDEX

Explanations

research papers

np_max-act · gemini-2.0-flash

The neuron activates on subword pieces of hyphenated technical terms and measurements (e.g. “three-dimensional,” “signal-to-noise,” “proof-of-work,” “contrast-enhanced”).

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

gap

-0.06

 [...]

-0.06

iyoruz

-0.06

 đào

-0.06

Combat

-0.06

่ใช

-0.06

getPost

-0.06

 blossom

-0.06

￣￣￣

-0.06

_mA

-0.06

POSITIVE LOGITS

특

0.06

ه

0.06

>--}}↵

0.06

.AllowUser

0.06

cq

0.06

(abs

0.06

of

0.06

_inv

0.06

 Glenn

0.06

Activations Density 0.018%

research papers

The neuron activates on subword pieces of hyphenated technical terms and measurements (e.g. “three-dimensional,” “signal-to-noise,” “proof-of-work,” “contrast-enhanced”).

No Comments

No Known Activations