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

0:0.10

1:0.08

2:0.05

3:0.08

4:0.09

5:0.08

6:0.07

7:0.09

8:0.09

9:0.08

10:0.07

11:0.07

Negative Logits

 cred

-2.25

aughter

-2.17

 Appl

-2.13

uments

-2.12

ationally

-2.03

imeo

-2.03

pport

-1.99

 sympath

-1.98

 unwitting

-1.98

 Plaintiff

-1.97

POSITIVE LOGITS

 anymore

2.11

ーク

1.98

ゼウス

1.97

Rot

1.95

 endings

1.95

etsk

1.94

 Breath

1.90

switch

1.89

 destruct

1.83

Choose

1.77

Activations Density 0.000%

No Known Activations

This feature has no known activations.