INDEX

Explanations

quantitative comparisons within scientific studies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

Whilst

-0.85

 Whilst

-0.81

 whilst

-0.77

 utilising

-0.74

‘

-0.74

 utilise

-0.69

 utilised

-0.67

Tikang

-0.64

 amongst

-0.63

şi

-0.62

POSITIVE LOGITS

 pleaſure

0.75

Efq

0.74

 Monfieur

0.68

 auffi

0.68

 becauſe

0.67

 perſon

0.66

 nont

0.65

 ſmall

0.65

 Theſe

0.65

 religieuse

0.63

Activations Density 1.595%

quantitative comparisons within scientific studies

No Comments

No Known Activations