INDEX

Explanations

phrases indicating significant concepts or ideas that underline essays and arguments

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.02

1:0.01

2:0.09

3:0.12

4:0.30

5:0.01

6:0.09

7:0.12

8:0.03

9:0.04

10:0.04

11:0.08

Negative Logits

ynes

-1.88

 favorably

-1.47

ussen

-1.42

���

-1.39

ebin

-1.32

 interacted

-1.32

 hadn

-1.32

augh

-1.31

Nex

-1.31

 UNCLASSIFIED

-1.29

POSITIVE LOGITS

 mysteries

1.52

 paradox

1.50

 mosa

1.42

 Infinite

1.39

blance

1.33

EEK

1.32

 fatal

1.31

 individuality

1.30

 misery

1.30

 carnage

1.29

Activations Density 0.001%

phrases indicating significant concepts or ideas that underline essays and arguments

No Comments

No Known Activations