INDEX

Explanations

concepts related to variable characteristics and contexts

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

UnusedPrivate

-0.67

awtextra

-0.59

AlterField

-0.59

OGND

-0.59

setVerticalGroup

-0.59

 snippetHide

-0.58

 surla

-0.54

НИК

-0.53

enterOuterAlt

-0.52

BufferException

-0.51

POSITIVE LOGITS

 avoient

0.37

 varies

0.37

 depending

0.35

 depender

0.35

 depends

0.34

 depended

0.33

opo

0.33

 depende

0.33

 Depending

0.32

odb

0.32

Activations Density 0.092%

concepts related to variable characteristics and contexts

No Comments

No Known Activations