INDEX

Explanations

occurrences of end-of-turn indicators in a programming or scripting context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

hysema

-0.41

ніципалі

-0.39

 outings

-0.38

esgue

-0.38

MEC

-0.36

なり

-0.36

 hàn

-0.35

Credential

-0.35

 Theorie

-0.35

 Credentials

-0.35

POSITIVE LOGITS

 المعيارى

0.66

 Infórmanos

0.66

 ویکی‌پدی

0.60

astéroïdes

0.55

WriteTagHelper

0.55

ioutil

0.50

 Exacts

0.49

 ErrIntOverflow

0.49

Rujuakan

0.49

 oprot

0.48

Activations Density 0.280%

occurrences of end-of-turn indicators in a programming or scripting context

No Comments

No Known Activations