INDEX

Explanations

various forms of the word "function" and related concepts signaling functionality or roles

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

__(/*!

-0.68

__*/

-0.67

RTSC

-0.53

 ddelweddau

-0.53

ronyms

-0.51

 GenerationType

-0.51

期刊论文

-0.50

LEncoder

-0.49

mannian

-0.48

 مرئيه

-0.48

POSITIVE LOGITS

 purpose

0.77

 function

0.70

 functions

0.66

purpose

0.66

Purpose

0.65

 purposes

0.65

 Purpose

0.65

 Function

0.64

function

0.57

Function

0.56

Activations Density 0.066%

various forms of the word "function" and related concepts signaling functionality or roles

No Comments

No Known Activations