INDEX

Explanations

fragments of code or programming constructs related to application contexts and operations involving components

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

 Chwiliwch

-0.37

naph

-0.35

TagMode

-0.35

btree

-0.33

Abbott

-0.32

SAM

-0.32

 Abbott

-0.31

Euc

-0.31

 Numerade

-0.31

 limp

-0.31

POSITIVE LOGITS

ValueStyle

0.67

 للمعارف

0.56

AccessorTable

0.55

 Paglinawan

0.51

getDoctrine

0.50

oredCriteria

0.50

HostException

0.50

IsMutable

0.49

 виправивши

0.47

 desmotivaciones

0.47

Activations Density 0.118%

fragments of code or programming constructs related to application contexts and operations involving components

No Comments

No Known Activations