INDEX

Explanations

bracketed or grouped expressions in mathematical content

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

Trọng

-0.68

السكان

-0.64

 ever

-0.61

valently

-0.61

 getCount

-0.55

 Reich

-0.55

 reten

-0.54

 Italijani

-0.53

ánto

-0.53

Reich

-0.53

POSITIVE LOGITS

تقاوى

0.68

 CreateTagHelper

0.64

tagHelperRunner

0.60

 Koro

0.55

 intptr

0.54

 transfieras

0.54

 Diſ

0.52

 propOrder

0.52

 }}"></

0.52

theless

0.52

Activations Density 0.164%

bracketed or grouped expressions in mathematical content

No Comments

No Known Activations