INDEX

Explanations

terms related to writing and writers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

 ögon

-0.77

ists

-0.77

meniz

-0.77

zono

-0.75

FROM

-0.73

EndContext

-0.72

Datuak

-0.71

 soldati

-0.70

 "..\..\

-0.69

 Clapton

-0.68

POSITIVE LOGITS

writer

1.02

 Writer

0.95

Writers

0.95

 writer

0.91

Writer

0.91

 Writers

0.90

 writers

0.88

 WRITER

0.77

writers

0.75

 Romain

0.74

Activations Density 0.087%

terms related to writing and writers

No Comments

No Known Activations