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

1:0.08

2:0.09

3:0.09

4:0.07

5:0.07

6:0.08

7:0.07

8:0.08

9:0.08

10:0.08

11:0.09

Negative Logits

odes

-2.01

 Glac

-1.96

ptions

-1.90

yip

-1.88

osate

-1.87

agascar

-1.87

ansk

-1.85

skirts

-1.85

warts

-1.77

iets

-1.75

POSITIVE LOGITS

Hew

1.86

��

1.61

 urging

1.52

 critic

1.51

 Rowe

1.51

�

1.50

 pledged

1.47

 acknowledging

1.44

 suggesting

1.41

 scathing

1.40

Activations Density 0.000%

No Known Activations

This feature has no known activations.