INDEX

Explanations

Applications and preparation

np_max-act · gemini-2.0-flash

The neuron activates on the numbered‐list and step‐heading tokens when the assistant is giving a structured, step-by-step set of suggestions.

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

Asp

-0.06

 army

-0.06

_attributes

-0.06

ระบบ

-0.06

 unconscious

-0.06

_CNT

-0.06

lichen

-0.06

něné

-0.06

Press

-0.06

ire

-0.06

POSITIVE LOGITS

-multi

0.07

:t

0.07

.");↵↵

0.07

 ActivityCompat

0.07

第四

0.07

allest

0.07

(NUM

0.06

 pupper

0.06

inging

0.06

Activations Density 0.042%

Applications and preparation

The neuron activates on the numbered‐list and step‐heading tokens when the assistant is giving a structured, step-by-step set of suggestions.

No Comments

No Known Activations

Applications and preparation

The neuron activates on the numbered‐list and step‐heading tokens when the assistant is giving a structured, step-by-step set of suggestions.

No Comments

No Known Activations