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

0:0.07

1:0.07

2:0.08

3:0.09

4:0.08

5:0.08

6:0.07

7:0.08

8:0.08

9:0.10

10:0.07

11:0.08

Negative Logits

cients

-1.71

ía

-1.67

stown

-1.58

imore

-1.57

 bustling

-1.56

itar

-1.51

NetMessage

-1.47

geon

-1.47

 Ange

-1.46

erie

-1.45

POSITIVE LOGITS

 redist

1.56

FFFF

1.36

 Transfer

1.35

PLA

1.34

 misuse

1.32

 cheat

1.32

 beneficiary

1.32

 tarn

1.30

Transfer

1.30

Optional

1.29

Activations Density 0.000%

No Known Activations

This feature has no known activations.