INDEX

Explanations

phrases related to past experiences or historical context

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

 defaultstate

-0.81

ngrx

-0.69

зулта

-0.63

قایناق‌لار

-0.62

MergeFrom

-0.60

ussis

-0.59

AutoresizingMask

-0.59

PRNewswire

-0.59

 AssemblyCulture

-0.58

 елның

-0.58

POSITIVE LOGITS

 pernah

1.09

เคย

1.07

 lived

1.03

 spent

1.01

 worked

0.95

spent

0.93

 từng

0.90

 played

0.86

 formerly

0.86

 Spent

0.85

Activations Density 0.384%

phrases related to past experiences or historical context

No Comments

No Known Activations