INDEX

Explanations

references to specific individuals' names and their prominence in various professional contexts

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

tag helper components

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

 matar

-0.32

 both

-0.29

 horaire

-0.28

 things

-0.28

 industriel

-0.28

 which

-0.28

 perus

-0.27

Pa

-0.27

 Schw

-0.27

 little

-0.26

POSITIVE LOGITS

AddTagHelper

0.80

tagHelperRunner

0.79

ſelf

0.76

 arşivlendi

0.74

 فريبيس

0.73

 fashiola

0.72

TagMode

0.72

 ſeine

0.71

jspx

0.69

 ſein

0.69

Activations Density 0.001%

references to specific individuals' names and their prominence in various professional contexts

tag helper components

No Comments

No Known Activations