INDEX

Explanations

bold or strongly emphasized text

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

↵↵

-0.77

<eos>

-0.70

Glej

-0.70

con

-0.66

to

-0.65

in

-0.62

-0.60

ini

-0.60

\\

-0.60

bu

-0.59

POSITIVE LOGITS

 myſelf

1.36

 purpoſe

1.28

 pleaſure

1.28

 itſelf

1.23

 ſtate

1.21

 ſche

1.21

 Majefty

1.17

 themſelves

1.17

<em>

1.16

 ſmall

1.14

Activations Density 0.034%

bold or strongly emphasized text

No Comments

No Known Activations