INDEX

Explanations

phrases indicating ease or complications related to processes or actions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

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

1:0.02

2:0.10

3:0.39

4:0.06

5:0.04

6:0.02

7:0.04

8:0.05

9:0.08

10:0.08

11:0.06

Negative Logits

xit

-1.78

orum

-1.77

lez

-1.76

inburgh

-1.66

isode

-1.65

INGTON

-1.60

ican

-1.59

psons

-1.58

claw

-1.57

�

-1.57

POSITIVE LOGITS

 perce

1.87

 inherently

1.84

 technically

1.70

DERR

1.65

 notoriously

1.64

 nature

1.64

 poorly

1.63

 physically

1.63

thumbnails

1.57

 finite

1.54

Activations Density 0.894%

phrases indicating ease or complications related to processes or actions

No Comments

No Known Activations