INDEX

Explanations

phrases related to legal or judicial contexts

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

0:0.01

1:0.01

2:0.10

3:0.07

4:0.13

5:0.02

6:0.03

7:0.39

8:0.03

9:0.04

10:0.08

11:0.04

Negative Logits

Offline

-1.38

cens

-1.23

except

-1.22

eff

-1.18

 deterior

-1.18

 calmly

-1.15

kept

-1.12

��

-1.11

rity

-1.10

 tolerated

-1.09

POSITIVE LOGITS

irgin

1.35

ader

1.17

 Shad

1.15

DOI

1.13

itan

1.12

 sovere

1.12

 Ranger

1.10

 Horizons

1.09

 Franch

1.09

 Flavoring

1.09

Activations Density 0.585%

phrases related to legal or judicial contexts

No Comments

No Known Activations