INDEX

Explanations

expressions of return or retrieval actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

rungsseite

-0.69

=$?

-0.67

LookAnd

-0.65

AndEndTag

-0.65

 Där

-0.60

antaranya

-0.59

ukone

-0.59

 setOpen

-0.59

TagHelpers

-0.59

mals

-0.58

POSITIVE LOGITS

Обра

0.48

dig

0.44

 sacerdote

0.44

 TextInputType

0.44

randomUUID

0.43

ędziesz

0.43

0.41

Incident

0.40

conomie

0.40

Activations Density 0.056%

expressions of return or retrieval actions

No Comments

No Known Activations