INDEX

Explanations

phrases that assert beliefs or clarify facts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_10/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.10.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

 ſtate

-0.83

 pleaſure

-0.81

 cuffs

-0.78

 purpoſe

-0.77

 ſame

-0.76

 RSSSF

-0.76

Portály

-0.75

Yoh

-0.75

 raiſ

-0.75

 Weyl

-0.75

POSITIVE LOGITS

 being

1.07

 also

1.02

is

0.98

be

0.97

 simply

0.84

was

0.84

not

0.83

 going

0.82

 être

0.80

 becoming

0.78

Activations Density 0.321%

phrases that assert beliefs or clarify facts

No Comments

No Known Activations