INDEX

Explanations

ed

np_max-act · gemini-2.0-flash

The neuron primarily responds to words ending in the past‐tense/participle suffix “-ed.”

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

allenges

-0.07

 ribs

-0.06

 позд

-0.06

toi

-0.06

(let

-0.06

тур

-0.06

Psi

-0.06

 repeating

-0.06

bindung

-0.06

 thing

-0.06

POSITIVE LOGITS

 Cyril

0.07

Sa

0.06

 Cục

0.06

MDB

0.06

ريف

0.06

.ItemStack

0.06

 دون

0.06

::-

0.06

Arg

0.06

softmax

0.06

Activations Density 0.044%

ed

The neuron primarily responds to words ending in the past‐tense/participle suffix “-ed.”

No Comments

No Known Activations