INDEX

Explanations

terms associated with extremes and limitless potentials

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

0:0.14

1:0.03

2:0.01

3:0.12

4:0.26

5:0.05

6:0.04

7:0.03

8:0.13

9:0.09

10:0.01

11:0.03

Negative Logits

ovan

-2.26

ousands

-1.99

thood

-1.97

onduct

-1.93

ursions

-1.88

failed

-1.84

redited

-1.79

 affili

-1.78

terrorist

-1.77

successfully

-1.76

POSITIVE LOGITS

 Curve

2.52

 nowhere

2.16

 Woodward

1.91

 BELOW

1.89

Whe

1.89

 Marse

1.89

achy

1.87

Emer

1.86

$$$$

1.85

 Neurolog

1.83

Activations Density 0.096%

terms associated with extremes and limitless potentials

No Comments

No Known Activations