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

0:0.08

1:0.07

2:0.09

3:0.08

4:0.10

5:0.06

6:0.07

7:0.08

8:0.08

9:0.07

10:0.08

11:0.08

Negative Logits

anamo

-2.35

utterstock

-1.88

akespe

-1.85

intent

-1.79

ilitary

-1.78

obook

-1.75

oteric

-1.73

 Bezos

-1.72

ahon

-1.72

inav

-1.72

POSITIVE LOGITS

 seism

1.79

Legend

1.78

 cooperative

1.53

Spawn

1.46

Angelo

1.45

 visitation

1.44

asionally

1.40

gui

1.40

Court

1.40

 tallest

1.39

Activations Density 0.000%

No Known Activations

This feature has no known activations.