INDEX

Explanations

general descriptions and summaries of concepts

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

broadly or general terms

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

 myſelf

-0.49

 Monfieur

-0.46

 ſta

-0.44

 uſe

-0.43

 pattes

-0.39

 ſtand

-0.39

 themſelves

-0.38

 uſed

-0.38

 neceff

-0.36

audrait

-0.36

POSITIVE LOGITS

 CURIAM

0.70

#+#

0.63

GEBURTSDATUM

0.60

大致

0.54

 ujednoznacz

0.52

TagMode

0.51

 AssemblyTitle

0.50

 summary

0.50

 summaries

0.50

summary

0.49

Activations Density 0.275%

general descriptions and summaries of concepts

broadly or general terms

No Comments

No Known Activations