INDEX

Explanations

tokens representing affiliations or component parts of groups

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

0:0.02

1:0.02

2:0.12

3:0.08

4:0.06

5:0.03

6:0.12

7:0.21

8:0.03

9:0.05

10:0.16

11:0.04

Negative Logits

abal

-1.87

staking

-1.78

nar

-1.56

recent

-1.55

 hostage

-1.53

ibal

-1.42

ophile

-1.35

 experiments

-1.32

 frontman

-1.32

selection

-1.30

POSITIVE LOGITS

Grade

1.92

inkle

1.75

 Dwell

1.57

endment

1.55

)=(

1.52

 condensed

1.49

 Except

1.48

Month

1.46

版

1.45

 Format

1.44

Activations Density 0.000%

tokens representing affiliations or component parts of groups

No Comments

No Known Activations