INDEX

Explanations

No Explanations Found

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

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.06

1:0.05

2:0.08

3:0.09

4:0.08

5:0.07

6:0.08

7:0.07

8:0.09

9:0.10

10:0.11

11:0.07

Negative Logits

alde

-1.93

ERROR

-1.90

ネ

-1.80

RIP

-1.79

 Invalid

-1.77

�

-1.75

APD

-1.75

realDonaldTrump

-1.74

━

-1.69

etheless

-1.69

POSITIVE LOGITS

‎

1.74

�

1.74

illac

1.67

aceutical

1.66

ibaba

1.58

 licens

1.58

 summarizes

1.57

¶

1.56

 Catalyst

1.56

Tai

1.52

Activations Density 0.000%

No Known Activations

This feature has no known activations.