INDEX

Explanations

locations and actions associated with measurements or evaluations

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_L10_Hcat_z_lr1.20e-03_l11.30e+00_ds24576_bs4096_dc1.00e-05_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.10.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

0:0.20

1:0.07

2:0.08

3:0.06

4:0.05

5:0.03

6:0.09

7:0.09

8:0.04

9:0.04

10:0.15

11:0.06

Negative Logits

 mart

-3.50

MFT

-3.23

etta

-3.19

Ib

-3.17

 Coyotes

-3.13

 Tyrann

-3.06

Aur

-2.92

 Bundy

-2.83

 Draco

-2.81

 clen

-2.81

POSITIVE LOGITS

Sew

8.77

sew

6.05

 sewer

5.08

 sewage

4.05

 sewing

3.90

 Roose

3.56

 Chatt

3.44

Va

3.38

Osw

3.22

 Sparrow

3.20

Activations Density 0.000%

locations and actions associated with measurements or evaluations

No Comments

No Known Activations