INDEX

Explanations

words related to improving or enhancing understanding and systems

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_21/width_16k/average_l0_70

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 vogliono

-0.45

-0.42

ories

-0.41

あれば

-0.41

ptest

-0.41

 căr

-0.41

ържа

-0.40

 mesmas

-0.39

 aceea

-0.39

POSITIVE LOGITS

 existing

1.22

existing

1.11

 bestehende

1.09

Existing

1.06

 Existing

1.06

 EXISTING

1.02

ValueStyle

1.02

 bestaande

0.96

 istnie

0.95

 bestehenden

0.93

Activations Density 0.562%

words related to improving or enhancing understanding and systems

No Comments

No Known Activations