INDEX

Explanations

punctuation

np_max-act · gemini-2.0-flash

The neuron fires when the user asks “do you know X?” type questions—that is, turns where the user is querying the assistant’s familiarity with a particular named entity.

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

Box

-0.07

big

-0.07

 Kathleen

-0.06

 Saturday

-0.06

kwargs

-0.06

odd

-0.06

(Account

-0.06

lst

-0.06

TH

-0.06

Oh

-0.06

POSITIVE LOGITS

овер

0.07

 regimes

0.06

 disciplinary

0.06

isper

0.06

ENUM

0.06

-directed

0.06

 incid

0.06

erap

0.06

GRAPH

0.06

.FileInputStream

0.06

Activations Density 0.035%

punctuation

The neuron fires when the user asks “do you know X?” type questions—that is, turns where the user is querying the assistant’s familiarity with a particular named entity.

No Comments

No Known Activations