INDEX

Explanations

terms related to mathematical or structured representations, possibly in the context of node and edge diagrams

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

setof

-0.65

={({

-0.62

httphttps

-0.62

kenn

-0.60

Hozzáférés

-0.60

WEBPACK

-0.59

 &___

-0.57

原始内容存档于

-0.57

ագրություններ

-0.56

Geplaatst

-0.54

POSITIVE LOGITS

 neither

0.79

 both

0.70

 Neither

0.64

 Both

0.62

Neither

0.62

逆に

0.59

neither

0.57

 either

0.56

Both

0.56

Ambos

0.54

Activations Density 2.035%

terms related to mathematical or structured representations, possibly in the context of node and edge diagrams

No Comments

No Known Activations