INDEX

Explanations

programming advice

np_max-act · gemini-2.0-flash

The neuron fires on imperative, how-to-style instruction words (e.g. “Use,” “Change,” “Move”) that kick off implementation advice.

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

ducer

-0.07

وسط

-0.06

γι

-0.06

 Kemal

-0.06

）、

-0.06

 Stadium

-0.06

$str

-0.06

 tablesp

-0.06

.listeners

-0.06

 Ston

-0.06

POSITIVE LOGITS

<style

0.07

 mają

0.07

Scaled

0.06

igDecimal

0.06

ctrine

0.06

RELATED

0.06

ipzig

0.06

_ak

0.06

	date

0.06

 praž

0.06

Activations Density 0.057%

programming advice

The neuron fires on imperative, how-to-style instruction words (e.g. “Use,” “Change,” “Move”) that kick off implementation advice.

No Comments

No Known Activations