INDEX

Explanations

phrases expressing possession or experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

 itſelf

-1.17

 themſelves

-1.05

Efq

-1.04

 himſelf

-0.96

 Monfieur

-0.90

 referenties

-0.87

 שוליים

-0.86

 Theſe

-0.85

ſelves

-0.85

 Jefus

-0.85

POSITIVE LOGITS

1.23

0.95

my

0.93

My

0.84

 myself

0.84

My

0.79

私は

0.78

我在

0.74

我也

0.72

ıyorum

0.72

Activations Density 0.306%

phrases expressing possession or experiences

No Comments

No Known Activations