INDEX

Explanations

instances of the word "unlike" to highlight contrasts or differences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

 produisons

-0.53

 sanitaires

-0.53

mockMvc

-0.52

 تعدى

-0.52

ebra

-0.49

findpost

-0.49

 mellom

-0.48

 vélos

-0.46

 deadliest

-0.46

 menak

-0.46

POSITIVE LOGITS

 unlike

1.15

unlike

1.15

 Unlike

1.08

Unlike

1.07

 Like

0.80

Like

0.72

 LIKE

0.69

CloseOperation

0.66

 otomatig

0.65

therly

0.64

Activations Density 0.262%

instances of the word "unlike" to highlight contrasts or differences

No Comments

No Known Activations