INDEX

Explanations

formal language

np_max-act · gemini-2.0-flash

The neuron flags self-assessment hedging about the model’s language proficiency (e.g. statements that its ability “may vary depending on the complexity of the text and context”).

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

Negative Logits

기를

-0.06

udging

-0.05

 lunch

-0.05

 döndü

-0.05

oo

-0.05

lide

-0.05

 지난

-0.05

 '}';↵

-0.05

POSITIVE LOGITS

_DIFF

0.07

 calculated

0.07

letters

0.06

shall

0.06

 sandals

0.06

>>>(

0.06

 fulfil

0.06

بش

0.06

.pathname

0.06

Activations Density 0.018%

formal language

The neuron flags self-assessment hedging about the model’s language proficiency (e.g. statements that its ability “may vary depending on the complexity of the text and context”).

No Known Activations

formal language

The neuron flags self-assessment hedging about the model’s language proficiency (e.g. statements that its ability “may vary depending on the complexity of the text and context”).

No Known Activations