INDEX

Explanations

phrases related to significant changes or actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_16k/average_l0_105

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

 irré

-0.74

Jumping

-0.65

 jumping

-0.64

Falling

-0.61

 constamment

-0.60

 touching

-0.60

Datuak

-0.60

jumping

-0.57

 personlig

-0.57

 bailando

-0.56

POSITIVE LOGITS

 walk

1.09

 rise

1.07

 hike

1.04

 roll

1.02

 move

0.99

 climb

0.98

 turn

0.94

 drop

0.92

 visit

0.90

 push

0.90

Activations Density 0.522%

phrases related to significant changes or actions

No Comments

No Known Activations