INDEX

Explanations

news-related information such as incidents, investigations, statements, and sources

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L2_Hcat_z_lr1.20e-03_l11.00e+00_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v4.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.2.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

0:0.07

1:0.02

2:0.06

3:0.08

4:0.06

5:0.05

6:0.05

7:0.09

8:0.18

9:0.05

10:0.15

11:0.09

Negative Logits

apon

-1.16

peat

-1.12

ointed

-0.95

 reverted

-0.93

cig

-0.92

 subjug

-0.91

rontal

-0.90

venge

-0.90

 peacefully

-0.88

owl

-0.88

POSITIVE LOGITS

catentry

1.11

reporting

1.07

OPLE

0.99

 circumstance

0.95

Str

0.94

advertisement

0.93

availability

0.93

Provider

0.92

 contingency

0.91

 spokesperson

0.90

Activations Density 0.099%

news-related information such as incidents, investigations, statements, and sources

No Comments

No Known Activations