INDEX

Explanations

thinking

np_max-act · gemini-2.0-flash

The neuron flags subjective hedging or opinion phrases (e.g. “I think,” “I don’t think,” “at least I think so”).

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

@",

-0.07

bindings

-0.06

.fastjson

-0.06

Un

-0.06

Uh

-0.06

Om

-0.06

UPS

-0.06

REMOVE

-0.06

blockquote

-0.06

کار

-0.06

POSITIVE LOGITS

 만들

0.07

 ejec

0.07

 Ев

0.06

+</

0.06

제

0.06

 його

0.06

müş

0.06

러운

0.06

_translate

0.06

(trigger

0.06

Activations Density 0.020%

thinking

The neuron flags subjective hedging or opinion phrases (e.g. “I think,” “I don’t think,” “at least I think so”).

No Comments

No Known Activations