INDEX

Explanations

unexpected outcomes

np_max-act · gemini-2.0-flash

The neuron responds to question phrasing, especially the word “what” (and related question-structure tokens like “you” and “expect”).

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

 InterruptedException

-0.07

еп

-0.06

导致

-0.06

aptop

-0.06

菌

-0.06

제

-0.06

 کوچک

-0.06

 захід

-0.06

kü

-0.06

پ

-0.06

POSITIVE LOGITS

 Prim

0.06

-result

0.06

parm

0.06

sel

0.06

 billions

0.06

";}↵

0.06

aggio

0.06

undefined

0.06

 planning

0.06

 TERM

0.06

Activations Density 0.015%

unexpected outcomes

The neuron responds to question phrasing, especially the word “what” (and related question-structure tokens like “you” and “expect”).

No Comments

No Known Activations