INDEX

Explanations

programming-related output statements and formatting elements

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 ब्रेकडाउन

-0.71

aarrggbb

-0.65

 kaarangay

-0.59

tagHelperRunner

-0.59

CppMethod

-0.59

 queſta

-0.57

 Chwiliwch

-0.57

fjspx

-0.54

IsMutable

-0.54

TagMode

-0.54

POSITIVE LOGITS

 comparison

0.40

 comparaison

0.38

 comparing

0.36

 mean

0.33

ajari

0.33

 favorite

0.32

 compare

0.32

 favored

0.31

($

0.31

 Cheesecake

0.31

Activations Density 0.006%

programming-related output statements and formatting elements

No Comments

No Known Activations