INDEX

Explanations

terms related to geometric or spatial concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_10/width_16k/average_l0_77

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

enga

-0.46

STR

-0.46

 berat

-0.45

ándolos

-0.45

 épisodes

-0.44

 Length

-0.44

 épisode

-0.44

 episódio

-0.44

KEYCODE

-0.43

ciąż

-0.43

POSITIVE LOGITS

 combinations

0.74

 Combinations

0.74

 corners

0.73

 intersection

0.69

клопе

0.69

AccessorTable

0.68

 combination

0.67

 intersections

0.66

 combinaison

0.66

 Combination

0.66

Activations Density 0.871%

terms related to geometric or spatial concepts

No Comments

No Known Activations