INDEX

Explanations

we

np_max-act · gemini-2.0-flash

This neuron primarily detects first-person plural pronouns (e.g. “we” / Russian “мы”).

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

 Prize

-0.07

saving

-0.07

.log

-0.07

')->

-0.07

%%↵

-0.06

*);↵↵

-0.06

 Morav

-0.06

 rotating

-0.06

 videog

-0.06

.default

-0.06

POSITIVE LOGITS

 Chúng

0.07

 chúng

0.07

均

0.07

 tradi

0.07

 còn

0.07

MOST

0.06

dex

0.06

Interaction

0.06

rk

0.06

dür

0.06

Activations Density 0.076%

we

This neuron primarily detects first-person plural pronouns (e.g. “we” / Russian “мы”).

No Comments

No Known Activations