INDEX

Explanations

expressions of affection or positive sentiment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 17-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_17/width_16k/average_l0_77

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.17.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

-1.02

 Monfieur

-0.90

ſelves

-0.84

Datuak

-0.82

 Majefty

-0.82

Efq

-0.82

 AssemblyTitle

-0.81

tableFuture

-0.81

 myſelf

-0.80

ConstraintMaker

-0.80

POSITIVE LOGITS

how

0.66

the

0.65

 seeing

0.55

 hearing

0.54

so

0.51

to

0.51

lamb

0.48

 banget

0.48

 everything

0.48

it

0.47

Activations Density 0.062%

expressions of affection or positive sentiment

No Comments

No Known Activations