INDEX

Explanations

No Explanations Found

New Auto-Interp

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

0:0.10

1:0.08

2:0.07

3:0.08

4:0.09

5:0.08

6:0.08

7:0.08

8:0.07

9:0.08

10:0.07

11:0.07

Negative Logits

 Mellon

-1.77

��

-1.74

��

-1.58

 Trust

-1.58

 trust

-1.57

 Allows

-1.56

owler

-1.54

 Practice

-1.53

 trusts

-1.51

 associate

-1.50

POSITIVE LOGITS

 surf

1.82

rawdownloadcloneembedreportprint

1.75

fec

1.74

liquid

1.74

BuyableInstoreAndOnline

1.72

エル

1.64

/>

1.61

 feces

1.59

 blur

1.59

м

1.51

Activations Density 0.000%

No Known Activations

This feature has no known activations.