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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.08

1:0.10

2:0.08

3:0.09

4:0.07

5:0.06

6:0.07

7:0.06

8:0.08

9:0.08

10:0.07

11:0.09

Negative Logits

 Favor

-1.60

 appl

-1.60

mp

-1.58

 pods

-1.56

 sidel

-1.54

 inhibitor

-1.52

 leases

-1.51

 streamed

-1.51

 Renew

-1.50

 subscription

-1.50

POSITIVE LOGITS

raph

1.92

ⓘ

1.85

bil

1.74

 attRot

1.67

Constructed

1.65

EVA

1.63

zar

1.61

alion

1.61

 Colomb

1.60

================================================================

1.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.