INDEX

Explanations

reported speech

np_max-act · gemini-2.0-flash

This neuron fires on reporting or attribution language—verbs and phrases that introduce or describe statements, meetings, or official responses (e.g. “said,” “claimed,” “met,” “response”).

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

 massac

-0.08

 rent

-0.07

 mask

-0.07

 Mutual

-0.06

Mix

-0.06

 Stand

-0.06

 rental

-0.06

 Licensed

-0.06

 straw

-0.06

ّد

-0.06

POSITIVE LOGITS

 />)↵

0.07

飛

0.06

%">↵

0.06

);↵

0.06

fis

0.06

 července

0.06

 "
↵

0.06

...↵↵↵↵

0.06

()){

0.06

 Manafort

0.06

Activations Density 0.050%

reported speech

This neuron fires on reporting or attribution language—verbs and phrases that introduce or describe statements, meetings, or official responses (e.g. “said,” “claimed,” “met,” “response”).

No Comments

No Known Activations

reported speech

This neuron fires on reporting or attribution language—verbs and phrases that introduce or describe statements, meetings, or official responses (e.g. “said,” “claimed,” “met,” “response”).

No Comments

No Known Activations