INDEX

Explanations

references to figures and visual data representations

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

0:0.07

1:0.11

2:0.05

3:0.04

4:0.02

5:0.27

6:0.04

7:0.05

8:0.05

9:0.11

10:0.07

11:0.05

Negative Logits

orkshire

-1.91

nesday

-1.79

yip

-1.66

unicip

-1.64

vironment

-1.61

mber

-1.61

racuse

-1.60

uador

-1.59

nz

-1.58

hya

-1.55

POSITIVE LOGITS

Ez

1.33

 conc

1.32

 truths

1.27

 summar

1.25

 innocent

1.23

 reminders

1.21

 confirm

1.18

Obj

1.17

 Brennan

1.17

 certainty

1.16

Activations Density 0.003%

references to figures and visual data representations

No Comments

No Known Activations