INDEX

Explanations

references to television or TV-related content

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_L7_Hcat_z_lr1.20e-03_l11.10e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

float32

Hook Name

blocks.7.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

0:0.06

1:0.07

2:0.06

3:0.06

4:0.11

5:0.08

6:0.13

7:0.06

8:0.05

9:0.05

10:0.03

11:0.18

Negative Logits

"—

-3.16

•

-2.94

-2.93

)—

-2.64

fter

-2.62

—

-2.51

•

-2.47

.—

-2.43

—"

-2.43

©

-2.30

POSITIVE LOGITS

merce

3.02

��

2.61

etheless

2.60

 Monarch

2.54

uman

2.45

 sqor

2.44

ーティ

2.42

DragonMagazine

2.35

 cryptoc

2.29

 comr

2.23

Activations Density 0.000%

references to television or TV-related content

No Comments

No Known Activations