INDEX

Explanations

instances of the word "Don't"

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

0:0.05

1:0.07

2:0.08

3:0.07

4:0.11

5:0.07

6:0.07

7:0.12

8:0.07

9:0.07

10:0.09

11:0.08

Negative Logits

 precon

-1.57

 hover

-1.53

 Parables

-1.51

龍�

-1.50

 Juno

-1.48

 Kepler

-1.45

 permitting

-1.35

 deferred

-1.34

Launch

-1.34

 Vance

-1.33

POSITIVE LOGITS

icter

1.88

amn

1.81

opathic

1.78

iola

1.75

iasis

1.71

ibaba

1.67

bard

1.65

aceae

1.64

iety

1.63

aturdays

1.63

Activations Density 0.000%

instances of the word "Don't"

No Comments

No Known Activations