INDEX

Explanations

ID

np_max-act · gemini-2.0-flash

This neuron detects the instruction phrasing asking whether a summary is “factually consistent” with the document.

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

Av

-0.07

 Wheel

-0.06

온

-0.06

_Al

-0.06

Latin

-0.06

_ms

-0.06

ीएम

-0.06

_equals

-0.06

geist

-0.06

_records

-0.06

POSITIVE LOGITS

 sire

0.07

 이는

0.06

Reg

0.06

INT

0.06

 yPos

0.06

 mensagem

0.06

(dep

0.06

 بش

0.06

ت

0.06

ZH

0.06

Activations Density 0.001%

ID

This neuron detects the instruction phrasing asking whether a summary is “factually consistent” with the document.

No Comments

No Known Activations