INDEX

Explanations

step/number

np_max-act · gemini-2.0-flash

The neuron fires on numbered “step” markers in code comments (e.g. “step 1:”, “step 2.3047:”), i.e. it detects comment annotations of sequential steps.

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

 kennen

-0.07

ル

-0.07

 Κατηγορία

-0.06

 Strong

-0.06

 دولتی

-0.06

scp

-0.06

ルの

-0.06

 hooks

-0.06

 wings

-0.06

 Daniel

-0.06

POSITIVE LOGITS

>=

0.06

 بص

0.06

 pimp

0.06

 alignSelf

0.06

cs

0.06

uncia

0.06

(~(

0.06

Salary

0.06

{lng

0.06

periment

0.05

Activations Density 0.013%

step/number

The neuron fires on numbered “step” markers in code comments (e.g. “step 1:”, “step 2.3047:”), i.e. it detects comment annotations of sequential steps.

No Comments

No Known Activations