INDEX

Explanations

phrases or statements that convey strong assertions or objections

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_L9_Hcat_z_lr1.20e-03_l11.20e+00_ds24576_bs4096_dc1.00e-06_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.9.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.11

1:0.01

2:0.03

3:0.08

4:0.01

5:0.02

6:0.13

7:0.03

8:0.04

9:0.37

10:0.01

11:0.10

Negative Logits

aucuses

-3.43

Aph

-3.42

dash

-3.34

CSV

-3.33

Lav

-3.12

zech

-3.08

 Symb

-3.06

 Jugg

-3.04

 Dropbox

-3.02

 Jaguars

-2.99

POSITIVE LOGITS

Reilly

8.17

 Reilly

7.84

Fox

5.00

Fox

4.38

 Ross

4.03

Ry

3.99

yre

3.96

lr

3.94

 Hayes

3.92

 Wheeler

3.90

Activations Density 0.001%

phrases or statements that convey strong assertions or objections

No Comments

No Known Activations