INDEX

Explanations

second

np_max-act · gemini-2.0-flash

The neuron selectively activates on numeric performance figures—decimal-encoded stats and placement numbers (e.g. “second 3.5898…,” “third 1.1787…,” etc.) indicating athletes’ or drivers’ results.

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

lexport

-0.08

 aprove

-0.07

 humans

-0.07

_functions

-0.07

_intersection

-0.07

 Papers

-0.07

(argv

-0.06

.dictionary

-0.06

ErrorHandler

-0.06

_lambda

-0.06

POSITIVE LOGITS

BCH

0.07

otechn

0.06

 lever

0.06

Ü

0.06

 dairy

0.06

，再

0.06

.REG

0.06

 CONDITION

0.06

 viewPager

0.06

("----------------

0.06

Activations Density 0.010%

second

The neuron selectively activates on numeric performance figures—decimal-encoded stats and placement numbers (e.g. “second 3.5898…,” “third 1.1787…,” etc.) indicating athletes’ or drivers’ results.

No Comments

No Known Activations