INDEX

Explanations

instructions or help

np_max-act · gemini-2.0-flash

This neuron detects instruction/request verbs (imperative task words like "create", "write", "design", "teach", "make") that signal a user asking the model to perform a task.

oai_token-act-pair · gpt-5-mini Triggered by @vetterc0

New Auto-Interp

Configuration

andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_15/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.15.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ходим

-0.07

ゾ

-0.06

uth

-0.06

۲۳

-0.06

 acclaim

-0.06

ิง

-0.06

Vil

-0.06

/co

-0.06

 ارائه

-0.06

-only

-0.06

POSITIVE LOGITS

(units

0.07

.delay

0.07

.UnitTesting

0.07

Decimal

0.07

.DataBind

0.07

Sensitive

0.07

.putInt

0.07

_jwt

0.07

dav

0.06

_OVERRIDE

0.06

Activations Density 0.119%

instructions or help

This neuron detects instruction/request verbs (imperative task words like "create", "write", "design", "teach", "make") that signal a user asking the model to perform a task.

No Comments

No Known Activations