INDEX

Explanations

terms related to ease and difficulty

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

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

 المعيارى

-0.59

centralwidget

-0.59

rungsseite

-0.57

bootstrapcdn

-0.52

StoryboardSegue

-0.51

 propOrder

-0.51

 HasFactory

-0.49

 Chwiliwch

-0.49

SuspendLayout

-0.48

 EconPapers

-0.47

POSITIVE LOGITS

不易

0.51

不容易

0.44

日至

0.36

IsContent

0.34

цезда

0.34

 stable

0.34

 odeur

0.34

 reten

0.34

 resistant

0.33

extAlignment

0.33

Activations Density 0.005%

terms related to ease and difficulty

No Comments

No Known Activations