INDEX

Explanations

terms related to the classification or grouping of items or concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

 greateſt

-0.81

 Theſe

-0.77

 AssemblyTitle

-0.72

 Diſ

-0.71

 Monfieur

-0.70

Eſ

-0.69

 poffible

-0.69

 Houſe

-0.68

 ويكيپيديا

-0.67

 Conſ

-0.67

POSITIVE LOGITS

 đều

1.19

ล้ว

0.66

均

0.64

皆

0.62

都

0.61

 both

0.61

都是

0.60

 CreateTagHelper

0.59

都有

0.58

All

0.56

Activations Density 0.182%

terms related to the classification or grouping of items or concepts

No Comments

No Known Activations