INDEX

Explanations

terms related to evaluation and comparison in descriptions and arguments

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

WithIOException

-0.53

 nakalista

-0.45

AndroidJUnit

-0.45

 defaultstate

-0.44

 IFTT

-0.43

 informée

-0.43

 Chwiliwch

-0.42

ectady

-0.42

 spermato

-0.41

TableBody

-0.40

POSITIVE LOGITS

atschappij

0.43

WithMany

0.42

lagi

0.40

 Wikimedijinoj

0.39

 élas

0.39

principalTable

0.38

opcional

0.38

衅

0.38

ARDO

0.38

 Girl

0.38

Activations Density 0.013%

terms related to evaluation and comparison in descriptions and arguments

No Comments

No Known Activations