INDEX

Explanations

of

np_max-act · gemini-2.0-flash

discussions on experimental methodologies in scientific research.

oai_token-act-pair · gpt-4o-mini Triggered by @xinyanhu8

This neuron is effectively “dead”—it never activates on any input and thereby detects no feature.

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

пов

-0.07

 Keith

-0.06

dependent

-0.06

 อย

-0.06

 Grove

-0.06

 ремон

-0.06

 Firewall

-0.06

xBE

-0.06

_compat

-0.06

 دش

-0.06

POSITIVE LOGITS

rud

0.06

[msg

0.06

 unlock

0.06

<?>>

0.06

GP

0.06

.executeQuery

0.06

 jusqu

0.06

 *);↵↵

0.06

.series

0.06

Activations Density 0.029%

of

discussions on experimental methodologies in scientific research.

This neuron is effectively “dead”—it never activates on any input and thereby detects no feature.

No Comments

No Known Activations