INDEX

Explanations

concepts related to simplification and ease of understanding or process improvement

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

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L2_Hcat_z_lr1.20e-03_l11.00e+00_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v4.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.2.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.09

1:0.05

2:0.12

3:0.10

4:0.11

5:0.05

6:0.03

7:0.03

8:0.15

9:0.14

10:0.06

11:0.02

Negative Logits

��

-1.38

acid

-1.18

 Nath

-1.15

weet

-1.14

pressed

-1.09

Alert

-1.08

 cele

-1.08

 Jeremiah

-1.08

Rare

-1.08

 mosqu

-1.08

POSITIVE LOGITS

 workflow

1.47

 procedure

1.40

 procedures

1.39

 approach

1.26

 processes

1.22

 scenario

1.21

GST

1.20

 grids

1.20

 process

1.18

 calculation

1.17

Activations Density 0.007%

concepts related to simplification and ease of understanding or process improvement

No Comments

No Known Activations