INDEX

Explanations

results related to health studies or trials

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

higher or lower levels

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 EconPapers

-0.66

 AssemblyCulture

-0.60

RTDA

-0.52

WithIOException

-0.51

WebMethod

-0.51

 GenerationType

-0.50

Джерела

-0.50



-0.48

 hObject

-0.47

 ioutil

-0.46

POSITIVE LOGITS

parametrize

0.42

 lack

0.41

 lacked

0.36

Hentet

0.36

assertj

0.36

 ciddi

0.35

 Khusus

0.33

readFile

0.33

 européennes

0.33

 négatif

0.32

Activations Density 0.264%

results related to health studies or trials

higher or lower levels

No Comments

No Known Activations