INDEX

Explanations

news articles

np_max-act · gemini-2.0-flash

The neuron activates on discourse‐level connective words (e.g. contrastive or causal transition markers like “despite,” “results,” “therefore,” etc.).

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

ekt

-0.07

 kinds

-0.07

 wann

-0.06

WCS

-0.06

 nights

-0.06

 اغ

-0.06

-income

-0.06

 systems

-0.06

ibration

-0.06

 kind

-0.06

POSITIVE LOGITS

 knih

0.06

 reinforcements

0.06

>",

0.06

):

0.06

ـ

0.06

발

0.06

BAL

0.06

//_

0.06

ruž

0.06

يكا

0.06

Activations Density 0.090%

news articles

The neuron activates on discourse‐level connective words (e.g. contrastive or causal transition markers like “despite,” “results,” “therefore,” etc.).

No Comments

No Known Activations