INDEX

Explanations

references to academic publications and citations, particularly in the context of scientific research

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 34-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_34/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.34.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

件事情

-0.62

 Incorporated

-0.60

 television

-0.58

 جدًا

-0.58

 miniaturka

-0.57

 onely

-0.57

 maksymal

-0.57

 beziehungs

-0.55

 Corporation

-0.54

 administration

-0.53

POSITIVE LOGITS

 Univ

0.72

 Assn

0.69

 probs

0.68

istoitu

0.67

ppl

0.67

 Natl

0.66

Univ

0.66

yrs

0.63

EndInit

0.63

Twp

0.62

Activations Density 2.547%

references to academic publications and citations, particularly in the context of scientific research

No Comments

No Known Activations