INDEX

Explanations

#

np_max-act · gemini-2.0-flash

The neuron fires on structural markers and formatting tokens—section numbers, hash marks, separator lines, and other layout cues indicating document structure.

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

Gran

-0.07

(dy

-0.07

Sey

-0.06

 xmax

-0.06

ёр

-0.06

_printf

-0.06

 Debug

-0.06

_GPIO

-0.06

ly

-0.06

 Guitar

-0.06

POSITIVE LOGITS

 coherence

0.07

 hierarchy

0.06

 detailing

0.06

 σει

0.06

 messy

0.06

 sebagai

0.06

 premise

0.06

 holders

0.06

creenshot

0.06

持续

0.06

Activations Density 0.002%

#

The neuron fires on structural markers and formatting tokens—section numbers, hash marks, separator lines, and other layout cues indicating document structure.

No Comments

No Known Activations

#

The neuron fires on structural markers and formatting tokens—section numbers, hash marks, separator lines, and other layout cues indicating document structure.

No Comments

No Known Activations