INDEX

Explanations

crimes and conspiracies

np_max-act · gemini-2.0-flash

The neuron selectively activates on past‐tense verb forms (e.g. “uncovered,” “hit,” “submitted”).

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

think

-0.07

SP

-0.06

 paralysis

-0.06

top

-0.06

 Refresh

-0.06

 csak

-0.06

academic

-0.06

 slides

-0.06

 bluetooth

-0.06

addr

-0.06

POSITIVE LOGITS

.WaitFor

0.07

Mut

0.07

 Krish

0.07

{{$

0.06

 rejuven

0.06

 обмеж

0.06

 почему

0.06

ADD

0.06

uyo

0.06

Activations Density 0.054%

crimes and conspiracies

The neuron selectively activates on past‐tense verb forms (e.g. “uncovered,” “hit,” “submitted”).

No Comments

No Known Activations