INDEX

Explanations

phrases related to reasoning or conclusions

oai_token-act-pair · gpt-3.5-turbo 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.05

1:0.03

2:0.15

3:0.06

4:0.27

5:0.04

6:0.03

7:0.03

8:0.13

9:0.09

10:0.06

11:0.02

Negative Logits

 lobb

-1.54

nesty

-1.33

vertisement

-1.32

 lobby

-1.29

zilla

-1.28

 breat

-1.26

aturdays

-1.24

ilyn

-1.24

 breathe

-1.24

imposed

-1.23

POSITIVE LOGITS

��

1.54

��

1.44

��

1.43

�

1.40

CHAT

1.35

��

1.34

NEC

1.28

龍契士

1.26

��

1.26

 Virtue

1.25

Activations Density 0.006%

phrases related to reasoning or conclusions

No Comments

No Known Activations