INDEX

Explanations

Contact information

np_max-act · gemini-2.0-flash

The neuron selectively activates on multi‐word proper names—especially organization or unit names (e.g. “Major Crime Unit,” “Provider Maintenance Team,” “Stony Brook Wildlife Sanctuary,” etc.).

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

pectral

-0.08

_version

-0.06

ğan

-0.06

andal

-0.06

_out

-0.06

卡

-0.06

 ту

-0.06

occasion

-0.06

out

-0.06

_Out

-0.06

POSITIVE LOGITS

饮

0.07

blast

0.07

.Mockito

0.07

 зрозум

0.07

 Vanessa

0.06

κολ

0.06

 експ

0.06

 Sergeant

0.06

 denying

0.06

 área

0.06

Activations Density 0.029%

Contact information

The neuron selectively activates on multi‐word proper names—especially organization or unit names (e.g. “Major Crime Unit,” “Provider Maintenance Team,” “Stony Brook Wildlife Sanctuary,” etc.).

No Comments

No Known Activations