INDEX

Explanations

phrases that suggest providing instructions or guidance

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

0:0.03

1:0.02

2:0.08

3:0.14

4:0.27

5:0.04

6:0.05

7:0.09

8:0.04

9:0.05

10:0.06

11:0.08

Negative Logits

龍

-1.75

amine

-1.47

..."

-1.45

ヘ

-1.42

OVA

-1.41

payer

-1.40

ATES

-1.40

soType

-1.40

 contingency

-1.36

YY

-1.33

POSITIVE LOGITS

interstitial

1.56

 sarc

1.41

 Gleaming

1.33

 indict

1.33

ntil

1.31

 citation

1.23

=>

1.22

 paraph

1.18

 invitations

1.17

 expulsion

1.16

Activations Density 0.000%

phrases that suggest providing instructions or guidance

No Comments

No Known Activations