INDEX

Explanations

recess

np_max-act · gemini-2.0-flash

The neuron activates on mentions of recessed or hollowed‐out features (tokens like “recess”/“recessed”).

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

 Bihar

-0.07

-export

-0.06

 вибор

-0.06

print

-0.06

ї

-0.06

ombine

-0.06

 타이

-0.06

-key

-0.06

你

-0.06

 tire

-0.06

POSITIVE LOGITS

 recess

0.10

 recession

0.09

CSC

0.08

Inset

0.08

 regress

0.07

 retreat

0.07

 دس

0.07

 recruiting

0.07

ako

0.07

 achievements

0.07

Activations Density 0.007%

recess

The neuron activates on mentions of recessed or hollowed‐out features (tokens like “recess”/“recessed”).

No Comments

No Known Activations

recess

The neuron activates on mentions of recessed or hollowed‐out features (tokens like “recess”/“recessed”).

No Comments

No Known Activations