INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L7_Hcat_z_lr1.20e-03_l11.10e+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.7.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

0:0.08

1:0.07

2:0.06

3:0.09

4:0.08

5:0.08

6:0.08

7:0.08

8:0.06

9:0.08

10:0.09

11:0.09

Negative Logits

Direct

-2.54

eus

-2.44

Expl

-2.42

ileged

-2.34

Credit

-2.34

Catalog

-2.33

 anytime

-2.26

=~=~

-2.25

itous

-2.17

Psy

-2.09

POSITIVE LOGITS

dylib

2.51

 Integrity

2.21

 lobbying

2.00

 hospitals

1.98

 Kirin

1.96

 actu

1.95

 Leilan

1.94

 Idle

1.90

 singer

1.90

 wrestler

1.87

Activations Density 0.000%

No Known Activations

This feature has no known activations.