INDEX

Explanations

phrases indicative of guidance or advice in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 pleaſure

-0.96

 houſe

-0.90

 itſelf

-0.90

ſelf

-0.89

 himſelf

-0.85

 poffible

-0.84

 Jefus

-0.84

findpost

-0.84

setVerticalGroup

-0.82

 purpoſe

-0.82

POSITIVE LOGITS

 fucked

0.68

...

0.67

ça

0.66

…

0.66

 nice

0.66

 haha

0.65

 shitty

0.65

 stupidly

0.65

 fucking

0.63

 shit

0.62

Activations Density 0.030%

phrases indicative of guidance or advice in various contexts

No Comments

No Known Activations