INDEX

Explanations

phrases related to structured formats or procedural guidelines

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

tvguidetime

-0.59

Efq

-0.55

 Lewin

-0.54

InitVars

-0.51

 <=",

-0.51

HttpPut

-0.50

 kanyang

-0.50

 hitherto

-0.50

 ''),

-0.50

 organiser

-0.49

POSITIVE LOGITS

AddHtmlAttribute

0.57

tanooga

0.55

itize

0.55

 typelib

0.55

kmäler

0.54

NSYLVANIA

0.54

Билгалдахарш

0.54

 Commencez

0.53

 виправивши

0.53

 AssemblyCompany

0.53

Activations Density 0.042%

phrases related to structured formats or procedural guidelines

No Comments

No Known Activations