INDEX

Explanations

programming problems and solutions

np_max-act-logits · gemini-2.0-flash

words indicating whether a technical solution works or successfully solves a problem.

oai_token-act-pair · claude-4-5-haiku Triggered by @xiaoqingsun004

New Auto-Interp

Configuration

andyrdt/saes-qwen2.5-7b-instruct/resid_post_layer_19/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.19.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

靛

-0.08

/,↵

-0.08

wn

-0.07

yö

-0.07

 çağ

-0.07

体验

-0.07

滗

-0.07

CTIONS

-0.07

 confidence

-0.07

튕

-0.06

POSITIVE LOGITS

电商

0.07

USE

0.07

 Numerous

0.07

地带

0.06

ouse

0.06

 breach

0.06

런

0.06

_:

0.06

Most

0.06

 Frauen

0.06

Activations Density 0.220%

programming problems and solutions

words indicating whether a technical solution works or successfully solves a problem.

No Comments

No Known Activations