INDEX

Explanations

prepositions and their associated contexts relating to location and direction

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

gebung

-0.32

går

-0.29

cznie

-0.29

Ārējās

-0.28

CM

-0.28

 jugó

-0.28

oare

-0.28

char

-0.27

break

-0.27

Level

-0.27

POSITIVE LOGITS

 BEHIND

0.91

 behind

0.85

Behind

0.85

behind

0.82

 Behind

0.82

 derrière

0.75

 onBackPressed

0.71

MessageTagHelper

0.71

 BACK

0.69

 dietro

0.69

Activations Density 0.026%

prepositions and their associated contexts relating to location and direction

No Comments

No Known Activations