INDEX

Explanations

specific terms related to medical applications and processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_21/width_16k/average_l0_70

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 Majefty

-0.94

 Chriftian

-0.86

 tryna

-0.85

 fubject

-0.82

 purpoſe

-0.81

 poffible

-0.79

 ftate

-0.77

 Deshalb

-0.76

 thingy

-0.76

 ſtate

-0.75

POSITIVE LOGITS

 esche

0.79

 Perhaps

0.62

 comuniques

0.61

 delà

0.61

Perhaps

0.60

 insure

0.59

ものである

0.59

perhaps

0.58

kwds

0.58

 perhaps

0.58

Activations Density 0.720%

specific terms related to medical applications and processes

No Comments

No Known Activations