INDEX

Explanations

goal

np_max-act · gemini-2.0-flash

The neuron fires on descriptive verbs and terms in technical/manual explanations (e.g. “shows,” “including,” “uses,” “size,” “invocations”).

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

 улы

-0.07

 kendisine

-0.06

(video

-0.06

/world

-0.06

 sexy

-0.06

	router

-0.06

 Только

-0.06

.truth

-0.06

�

-0.06

enant

-0.06

POSITIVE LOGITS

πο

0.08

_BEGIN

0.07

 München

0.07

_pieces

0.06

 ресурс

0.06

.DataFrame

0.06

 inconsistencies

0.06

ustrial

0.06

_RX

0.06

Activations Density 0.014%

goal

The neuron fires on descriptive verbs and terms in technical/manual explanations (e.g. “shows,” “including,” “uses,” “size,” “invocations”).

No Comments

No Known Activations