INDEX

Explanations

phrases indicating tendencies or inclinations

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

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.03

1:0.02

2:0.14

3:0.03

4:0.31

5:0.07

6:0.03

7:0.02

8:0.11

9:0.12

10:0.05

11:0.02

Negative Logits

Bay

-1.40

asp

-1.38

 redacted

-1.37

RAW

-1.36

country

-1.35

 Bounty

-1.32

fleet

-1.32

Eye

-1.32

ighth

-1.32

 helicop

-1.29

POSITIVE LOGITS

iru

1.61

 overest

1.51

 misinterpret

1.50

 misunderstand

1.49

 underestimate

1.48

 miscon

1.44

itual

1.43

 disson

1.40

omorphic

1.39

 stereotype

1.37

Activations Density 0.007%

phrases indicating tendencies or inclinations

No Comments

No Known Activations