INDEX

Explanations

gibberish and code

np_max-act · gemini-2.0-flash

This neuron detects tokens that are part of technical document or drawing references (e.g. the word “Drawing” and related specification identifiers).

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

 ende

-0.07

'>

-0.06

geries

-0.06

danger

-0.06

디

-0.06

 Kurulu

-0.06

phem

-0.06

 biomass

-0.06

 BOOK

-0.06

 журн

-0.06

POSITIVE LOGITS

_MISSING

0.08

 εισ

0.07

 доз

0.07

 favorable

0.07

_MINOR

0.06

 Basketball

0.06

 violently

0.06

激

0.06

_agent

0.06

Activations Density 0.000%

gibberish and code

This neuron detects tokens that are part of technical document or drawing references (e.g. the word “Drawing” and related specification identifiers).

No Comments

No Known Activations