INDEX

Explanations

phrases indicating a basis or foundation for arguments or claims

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

RegressionTest

-0.51

 myſelf

-0.50

MigrationBuilder

-0.49

 pleaſure

-0.46

 houſe

-0.46

ſy

-0.45

Interviewer

-0.43

 unforgettable

-0.42

logger

-0.42

 niemand

-0.41

POSITIVE LOGITS

 based

1.01

 BASED

0.96

based

0.95

 Based

0.93

Based

0.88

BASED

0.88

 dựa

0.82

 basado

0.81

 basados

0.78

 basé

0.75

Activations Density 0.123%

phrases indicating a basis or foundation for arguments or claims

No Comments

No Known Activations