INDEX

Explanations

terms related to health assessments and their complexities

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_L10_Hcat_z_lr1.20e-03_l11.30e+00_ds24576_bs4096_dc1.00e-05_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.10.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

0:0.07

1:0.06

2:0.06

3:0.05

4:0.07

5:0.03

6:0.22

7:0.14

8:0.02

9:0.04

10:0.06

11:0.11

Negative Logits

 hasht

-2.47

 radicals

-2.43

 retweet

-2.42

weet

-2.40

 livestream

-2.30

 hashtag

-2.28

 riot

-2.26

 flavours

-2.25

 collaborations

-2.21

 revolutions

-2.19

POSITIVE LOGITS

 assessment

5.13

 Assessment

4.94

 assessments

4.92

 evaluation

4.65

 assess

4.48

 Evaluation

4.48

 assessed

4.47

 appraisal

4.31

 evaluations

4.21

 eval

4.05

Activations Density 0.009%

terms related to health assessments and their complexities

No Comments

No Known Activations