INDEX

Explanations

active verbs related to ongoing actions or developments

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

0:0.03

1:0.03

2:0.25

3:0.03

4:0.02

5:0.04

6:0.15

7:0.06

8:0.11

9:0.12

10:0.07

11:0.03

Negative Logits

 regress

-1.15

 Codes

-1.10

RET

-1.07

 endings

-0.99

 observer

-0.96

 satisfaction

-0.94

 ranking

-0.94

 resignation

-0.92

 ethics

-0.91

 Fein

-0.91

POSITIVE LOGITS

myra

1.29

nels

1.25

merga

1.19

minecraft

1.19

aband

1.18

leck

1.17

エル

1.16

enegger

1.15

cffff

1.15

qqa

1.14

Activations Density 0.035%

active verbs related to ongoing actions or developments

No Comments

No Known Activations