INDEX

Explanations

only

np_max-act · gemini-2.0-flash

This neuron fires on occurrences of the biconditional cue “if and only if,” i.e. the word “only” in logical equivalence statements.

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

>:

-0.07

Ascending

-0.06

に

-0.06

 proced

-0.06

ρει

-0.06

、大

-0.06

計

-0.06

Vys

-0.06

_Template

-0.06

/board

-0.06

POSITIVE LOGITS

iff

0.08

pořád

0.07

 nevě

0.07

řád

0.06

leftrightarrow

0.06

khi

0.06

Bg

0.06

haust

0.06

 breadth

0.06

lever

0.06

Activations Density 0.004%

only

This neuron fires on occurrences of the biconditional cue “if and only if,” i.e. the word “only” in logical equivalence statements.

No Comments

No Known Activations