INDEX

Explanations

specific patterns or repetition of the token "le" followed by various sequences, indicating a focus on particular sequences involving "le"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_11/width_16k/average_l0_80

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

findpost

-0.92

eabouts

-0.86

DotNetBar

-0.80

 Herrick

-0.78

/*

-0.77

guous

-0.74

 Anastasia

-0.72

--}}

-0.70

 Newsroom

-0.70

 meatballs

-0.70

POSITIVE LOGITS

LE

1.19

LE

1.18

le

1.16

 Leighton

1.11

 Leona

1.10

 Leicestershire

1.05

Le

1.05

 Leandro

1.03

Vle

0.97

 Leal

0.97

Activations Density 0.193%

specific patterns or repetition of the token "le" followed by various sequences, indicating a focus on particular sequences involving "le"

No Comments

No Known Activations