INDEX

Explanations

code or programming

np_max-act · gemini-2.0-flash

This neuron fires on occurrences of the word “The” when it appears at the start of high‐level specification or intent statements (e.g. “The intention of the program is…”).

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

 aromatic

-0.06

.nb

-0.06

phinx

-0.06

 překlad

-0.06

explode

-0.06

 promotes

-0.06

peak

-0.06

cmd

-0.06

PERTIES

-0.06

POSITIVE LOGITS

shal

0.07

 sugars

0.07

avirus

0.06

ARS

0.06

 missiles

0.06

дия

0.06

 rout

0.06

 tắc

0.06

 solar

0.06

 sieve

0.06

Activations Density 0.009%

code or programming

This neuron fires on occurrences of the word “The” when it appears at the start of high‐level specification or intent statements (e.g. “The intention of the program is…”).

No Comments

No Known Activations