INDEX

Explanations

references to various types of evidence and their reliability in decision making

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

evidence and descriptions

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

AndEndTag

-0.41

exp

-0.35

 καν

-0.29

зулта

-0.27

 finally

-0.27

 only

-0.27

 easiest

-0.27

oul

-0.26

 wild

-0.26

 parem

-0.26

POSITIVE LOGITS

 initComponents

0.61

styleType

0.59

ArrowToggle

0.58

ThroughAttribute

0.57

requireNonNull

0.56

enciaga

0.56

AddTagHelper

0.56

Tembelea

0.54

awtextra

0.54

ErrUnexpectedEOF

0.54

Activations Density 0.082%

references to various types of evidence and their reliability in decision making

evidence and descriptions

No Comments

No Known Activations