INDEX

Explanations

usage of directional terms and references to 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_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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.09

1:0.02

2:0.03

3:0.19

4:0.04

5:0.05

6:0.06

7:0.04

8:0.04

9:0.01

10:0.35

11:0.03

Negative Logits

 refunds

-1.96

 autonomy

-1.93

 Innocent

-1.81

 warranty

-1.77

 infall

-1.76

 unlimited

-1.74

======

-1.71

 entirely

-1.67

necess

-1.66

profit

-1.66

POSITIVE LOGITS

 flanked

2.68

-)

2.21

aucus

2.21

/)

2.16

)'

2.15

finger

2.09

WARD

2.04

izont

2.04

behind

1.97

)",

1.95

Activations Density 0.016%

usage of directional terms and references to positions

No Comments

No Known Activations