INDEX

Explanations

references to news agencies and media reporting

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

 íntimo

-0.55

printLine

-0.46

 MotionEvent

-0.44

 برانيه

-0.43

createServer

-0.43

 Magazine

-0.41

 Zeitschrift

-0.40

Magazine

-0.38

Mag

-0.38

 مواليد

-0.37

POSITIVE LOGITS

 agency

1.07

 wire

0.97

 Agencies

0.96

 agencies

0.96

 Reuters

0.94

 Agency

0.94

swire

0.93

agency

0.91

 wires

0.90

wires

0.90

Activations Density 0.119%

references to news agencies and media reporting

No Comments

No Known Activations