INDEX

Explanations

code repository references

np_max-act · gemini-2.0-flash

This neuron activates on segments of badge or repository URLs (e.g. domain and path tokens like “github.com,” “coveralls.io,” “codecov.io,” “matn/go-isatty,” etc.)

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

 yapan

-0.06

 Marine

-0.06

ungeon

-0.06

 मह

-0.06

-et

-0.06

加

-0.06

 Mong

-0.06

iri

-0.06

du

-0.06

 Navy

-0.06

POSITIVE LOGITS

 Üst

0.07

ัจจ

0.06

되었습니다

0.06

瞬

0.06

lbrakk

0.06

ampling

0.06

(HttpContext

0.06

 Stateless

0.06

 wannonce

0.06

０

0.06

Activations Density 0.006%

code repository references

This neuron activates on segments of badge or repository URLs (e.g. domain and path tokens like “github.com,” “coveralls.io,” “codec­ov.io,” “matn/go-isatty,” etc.)

No Comments

No Known Activations

This neuron activates on segments of badge or repository URLs (e.g. domain and path tokens like “github.com,” “coveralls.io,” “codecov.io,” “matn/go-isatty,” etc.)