INDEX

Explanations

links and shares associated with 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_L5_Hcat_z_lr1.20e-03_l11.00e+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.5.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

1:0.12

2:0.08

3:0.08

4:0.04

5:0.11

6:0.08

7:0.01

8:0.10

9:0.09

10:0.05

11:0.08

Negative Logits

cffff

-1.74

ndum

-1.73

 largeDownload

-1.53

jug

-1.53

 ILCS

-1.52

ancial

-1.50

abwe

-1.50

escription

-1.49

atform

-1.48

 Aires

-1.48

POSITIVE LOGITS

Classic

1.68

 competitors

1.28

 Classic

1.25

 competitor

1.21

 favorites

1.17

Winner

1.16

1.15

morning

1.15

 users

1.15

Heart

1.13

Activations Density 0.000%

links and shares associated with content

No Comments

No Known Activations