INDEX

Explanations

research hypotheses

np_max-act · gemini-2.0-flash

This neuron activates on the hedging and predictive language used to state research hypotheses or aims (e.g. “we hypothesized that,” “would,” “aimed to,” “differences,” etc.).

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

cc

-0.07

 physicists

-0.07

нев

-0.07

 butterfly

-0.06

igos

-0.06

 Main

-0.06

delegate

-0.06

ادی

-0.06

journal

-0.06

anche

-0.06

POSITIVE LOGITS

userManager

0.07

<center

0.07

 Finding

0.07

 GetLastError

0.07

 أك

0.06

(emp

0.06

.Json

0.06

 เร

0.06

 přen

0.06

.XPATH

0.06

Activations Density 0.014%

research hypotheses

This neuron activates on the hedging and predictive language used to state research hypotheses or aims (e.g. “we hypothesized that,” “would,” “aimed to,” “differences,” etc.).

No Comments

No Known Activations