INDEX

Explanations

attends to the presence of empty tokens surrounded by asterisks from subsequent token references

oai_attention-head · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-att-16k

Configuration

google/gemma-scope-2b-pt-att/layer_10/width_16k/average_l0_70

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.attn.hook_z

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.08

1:0.09

2:0.08

3:0.16

4:0.16

5:0.08

6:0.20

7:0.11

Negative Logits

findpost

-0.46

GEBURTSDATUM

-0.41

 disambiguazione

-0.38



-0.36

ValueStyle

-0.36

nodoc

-0.35

Jegyzetek

-0.35

ConstraintMaker

-0.35

 AssemblyTitle

-0.33

queryInterface

-0.33

POSITIVE LOGITS

 grammi

0.30

 tây

0.28

 scho

0.28

 Schme

0.28

 '\\;'

0.27

 congi

0.26

Sơ

0.26

onday

0.26

Commun

0.26

 Sherwood

0.26

Activations Density 0.023%

attends to the presence of empty tokens surrounded by asterisks from subsequent token references

No Comments

No Known Activations