INDEX

Explanations

phrases that indicate departure or leaving

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

0:0.06

1:0.07

2:0.09

3:0.07

4:0.10

5:0.07

6:0.07

7:0.13

8:0.07

9:0.06

10:0.07

11:0.08

Negative Logits

 Legion

-1.60

 retaliation

-1.54

 herb

-1.49

 Ogre

-1.48

 flowering

-1.45

 weeds

-1.44

 Indra

-1.43

reath

-1.43

 Daylight

-1.42

 Adamant

-1.39

POSITIVE LOGITS

acebook

1.99

田

1.87

PLIED

1.85

ographics

1.82

 Parables

1.76

artments

1.75

 guiActiveUn

1.75

ypes

1.70

ophysical

1.67

ologist

1.67

Activations Density 0.000%

phrases that indicate departure or leaving

No Comments

No Known Activations