INDEX

Explanations

code/data sections

np_max-act · gemini-2.0-flash

This neuron detects the presence of a structured answer‐format template, marked by code‐block delimiters (```), a “Decision:” label, and bracketed fields.

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

Using

-0.07

 changes

-0.07

)n

-0.06

closed

-0.06

 komt

-0.06

‐'

-0.06

tou

-0.06

 concludes

-0.06

 touches

-0.06

_AP

-0.06

POSITIVE LOGITS

 зав

0.07

 зокрема

0.07

ούς

0.07

ctica

0.07

 сир

0.06

 наказ

0.06

adder

0.06

????

0.06

 olds

0.06

신

0.06

Activations Density 0.041%

code/data sections

This neuron detects the presence of a structured answer‐format template, marked by code‐block delimiters (```), a “Decision:” label, and bracketed fields.

No Comments

No Known Activations

code/data sections

This neuron detects the presence of a structured answer‐format template, marked by code‐block delimiters (```), a “Decision:” label, and bracketed fields.

No Comments

No Known Activations