INDEX

Explanations

instances of specific contextual signals or markers indicating a structured presentation of information

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_9/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.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

 ис

-0.52

ClCompile

-0.52

didSet

-0.51

kloped

-0.51

⊱

-0.50

parency

-0.49

Gek

-0.49

ExtendWith

-0.49

valbard

-0.49

Extinguishing

-0.49

POSITIVE LOGITS

tvguidetime

0.72

 AssemblyProduct

0.70

rungsseite

0.68

writeFieldEnd

0.64

'])->

0.63

 kone

0.62

 للاسماء

0.58

 GetEnumerator

0.58

 AssemblyCulture

0.58

]!='

0.56

Activations Density 0.082%

instances of specific contextual signals or markers indicating a structured presentation of information

No Comments

No Known Activations