INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L9_Hcat_z_lr1.20e-03_l11.20e+00_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.9.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

2:0.08

3:0.08

4:0.08

5:0.07

6:0.09

7:0.08

8:0.07

9:0.08

10:0.08

11:0.08

Negative Logits

aukee

-3.07

verson

-2.85

Anne

-2.85

elsen

-2.64

Dan

-2.62

 Clancy

-2.55

evin

-2.54

aron

-2.53

aylor

-2.53

Jen

-2.52

POSITIVE LOGITS

 Zeus

3.04

 snake

2.88

pus

2.88

 poisoning

2.73

 folly

2.70

XII

2.70

 Typh

2.69

ulhu

2.58

 nonsense

2.58

 plague

2.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.