INDEX

Explanations

references to the concept of awaiting or implying something in a given context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_3/width_16k/average_l0_90

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

 meaning

-0.67

 betekenis

-0.60

 Houſe

-0.59

 ſta

-0.58

 pouvoit

-0.55

 auroit

-0.52

 pleaſure

-0.52

 présidenti

-0.52

 ſtate

-0.52

tagez

-0.51

POSITIVE LOGITS

 imply

0.93

 implied

0.92

 implies

0.79

imply

0.71

 estekak

0.68

 IMPLIED

0.62

 implying

0.61

 implicitly

0.59

 await

0.57

 committed

0.56

Activations Density 0.138%

references to the concept of awaiting or implying something in a given context

No Comments

No Known Activations