INDEX

Explanations

terms related to various forms of classification or categorization, particularly in social or organizational contexts

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_L3_Hcat_z_lr1.20e-03_l19.00e-01_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.3.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

0:0.04

1:0.05

2:0.05

3:0.03

4:0.02

5:0.02

6:0.06

7:0.11

8:0.27

9:0.03

10:0.08

11:0.20

Negative Logits

quickShipAvailable

-1.56

{:

-1.29

 Ital

-1.27

 Orbital

-1.22

 Lust

-1.16

 ital

-1.15

 Genius

-1.13

 Pearce

-1.12

 Cobra

-1.12

 Philips

-1.09

POSITIVE LOGITS

ledged

1.08

icipated

1.07

iazep

1.06

 pickup

1.05

 sighting

1.05

 outing

1.03

ively

1.00

isable

1.00

ional

0.98

IZ

0.98

Activations Density 0.072%

terms related to various forms of classification or categorization, particularly in social or organizational contexts

No Comments

No Known Activations