INDEX

Explanations

references to evidence-based practices and mentions of specific political affiliations within health 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_L5_Hcat_z_lr1.20e-03_l11.00e+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.5.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

0:0.02

1:0.00

2:0.11

3:0.35

4:0.08

5:0.03

6:0.04

7:0.05

8:0.07

9:0.07

10:0.08

11:0.04

Negative Logits

furt

-1.71

 Leban

-1.51

 Pengu

-1.51

 Hanson

-1.47

}}}

-1.47

anooga

-1.46

VERTISEMENT

-1.46

peria

-1.44

berus

-1.43

 Nieto

-1.42

POSITIVE LOGITS

 "…

1.68

 ALWAYS

1.57

 oxid

1.55

 REALLY

1.53

("

1.52

Firstly

1.50

1.47

 distilled

1.46

(=

1.43

=(

1.42

Activations Density 0.020%

references to evidence-based practices and mentions of specific political affiliations within health contexts

No Comments

No Known Activations