INDEX

Explanations

Instructions

np_max-act · gemini-2.0-flash

The neuron fires on wording that signals completion of or transition between tutorial steps—phrases like “once you have…,” “added,” “chosen,” or “integrated” in instructional how-to explanations.

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

 Bright

-0.07

rn

-0.07

 Biggest

-0.07

uf

-0.07

messages

-0.07

alive

-0.07

 Therm

-0.06

Girl

-0.06

 tisíc

-0.06

 прип

-0.06

POSITIVE LOGITS

);;↵

0.06

BOOLE

0.06

_dims

0.06

 та

0.06

 carp

0.06

./

0.05

_payments

0.05

.setState

0.05

_sched

0.05

ъек

0.05

Activations Density 0.074%

Instructions

The neuron fires on wording that signals completion of or transition between tutorial steps—phrases like “once you have…,” “added,” “chosen,” or “integrated” in instructional how-to explanations.

No Comments

No Known Activations