INDEX

Explanations

scientific terminology related to evidence and explanations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 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

 Chwiliwch

-0.44

featureID

-0.40

 faſt

-0.40

MigrationBuilder

-0.39

BibitemShut

-0.39

culable

-0.38

 JpaRepository

-0.37

 posters

-0.37

kháu

-0.37

glBegin

-0.35

POSITIVE LOGITS

 hypothesis

0.61

hypothesis

0.53

 hypothesized

0.52

 hipótesis

0.51

 explaining

0.50

 interpreting

0.50

 interpret

0.49

 asumi

0.49

 explanation

0.49

 Hypothesis

0.48

Activations Density 1.149%

scientific terminology related to evidence and explanations

No Comments

No Known Activations