INDEX

Explanations

single digit numbers

np_max-act · gemini-2.0-flash

The neuron fires on programming‐language syntax (code snippets)—it identifies lines and tokens belonging to code blocks rather than natural‐language text.

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

IDE

-0.07

for

-0.07

 vertical

-0.06

licken

-0.06

ैट

-0.06

 Dark

-0.06

"For

-0.06

ored

-0.06

 curls

-0.06

Fade

-0.06

POSITIVE LOGITS

 sadness

0.07

 interess

0.07

 pardon

0.07

hf

0.07

 okres

0.06

 sorrow

0.06

noe

0.06

 artillery

0.06

.PO

0.06

 miglior

0.06

Activations Density 0.023%

single digit numbers

The neuron fires on programming‐language syntax (code snippets)—it identifies lines and tokens belonging to code blocks rather than natural‐language text.

No Comments

No Known Activations