INDEX

Explanations

names and terms related to geographic regions or entities

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

 للاسماء

-0.83

TemporalType

-0.77

featureID

-0.75

 dAtA

-0.75

intios

-0.72

+#+

-0.70

styleType

-0.70

<unused16>

-0.69

<unused14>

-0.69

[@BOS@]

-0.69

POSITIVE LOGITS

Kh

0.74

Kh

0.69

 Khal

0.63

kh

0.62

Khal

0.58

0.57

Kha

0.57

Kha

0.56

 Khat

0.54

 Khan

0.52

Activations Density 0.022%

names and terms related to geographic regions or entities

No Comments

No Known Activations