INDEX

Explanations

terms related to political claims and negotiations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 38-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_38/width_16k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.38.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

 zijne

-1.22

 dezelve

-1.15

 quæ

-1.12

 ſont

-1.08

 zoude

-1.05

 titolata

-1.03

 étoit

-1.03

 viață

-1.02

 avoient

-1.00

genodigd

-0.99

POSITIVE LOGITS

 recieve

0.71

 recieved

0.71

 seperate

0.69

com

0.69

 inte

0.68

 prode

0.68

 affor

0.67

 recie

0.66

 fash

0.66

 chall

0.66

Activations Density 2.933%

terms related to political claims and negotiations

No Comments

No Known Activations