INDEX

Explanations

phrases related to impact, consequences, and causation

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_L7_Hcat_z_lr1.20e-03_l11.10e+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.7.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

1:0.07

2:0.02

3:0.08

4:0.03

5:0.08

6:0.09

7:0.31

8:0.07

9:0.06

10:0.07

11:0.03

Negative Logits

ワン

-2.26

 Cosponsors

-2.10

quished

-2.08

Peb

-2.02

 Sitting

-2.00

 surn

-1.99

armed

-1.91

romeda

-1.90

 numbered

-1.89

 Beaver

-1.89

POSITIVE LOGITS

 enhances

3.19

 exacerb

3.17

 prevents

3.15

 reduces

3.09

 eliminates

2.94

 undermines

2.88

 aggrav

2.88

 allev

2.86

 improves

2.85

 inhibits

2.84

Activations Density 0.497%

phrases related to impact, consequences, and causation

No Comments

No Known Activations