INDEX

Explanations

references to software modules and their metadata

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

top level domains

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

 propOrder

-0.77

IsContent

-0.67

 Roskov

-0.66

 EconPapers

-0.65

ktop

-0.63

 InputDecoration

-0.62

 '\\;'

-0.61

 CreateTagHelper

-0.61

ValueStyle

-0.60

 gynhyrchwyd

-0.59

POSITIVE LOGITS

 Sigue

0.31

tvguidetime

0.28

 оригіналу

0.26

Efq

0.26

yet

0.26

ทุน

0.26

Yet

0.25

yet

0.25

YET

0.25

Hug

0.25

Activations Density 0.010%

references to software modules and their metadata

top level domains

No Comments

No Known Activations