INDEX

Explanations

phrases indicating retreat or retraction from statements or positions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.01

1:0.00

2:0.11

3:0.07

4:0.17

5:0.02

6:0.04

7:0.36

8:0.02

9:0.03

10:0.07

11:0.04

Negative Logits

 ILCS

-1.65

enery

-1.56

nesota

-1.51

 Ratings

-1.50

�

-1.50

lehem

-1.49

 Located

-1.48

among

-1.47

wow

-1.47

 Scores

-1.47

POSITIVE LOGITS

 indefinitely

1.95

 prematurely

1.88

 altogether

1.74

 temptation

1.73

 forcefully

1.73

 blatantly

1.71

 blindly

1.71

 insin

1.70

 delaying

1.69

 foolish

1.69

Activations Density 0.002%

phrases indicating retreat or retraction from statements or positions

No Comments

No Known Activations