INDEX

Explanations

the letter 'e'

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

0:0.13

1:0.03

2:0.03

3:0.10

4:0.04

5:0.09

6:0.10

7:0.02

8:0.26

9:0.08

10:0.03

11:0.04

Negative Logits

thood

-1.69

aughs

-1.56

emale

-1.52

iety

-1.49

peak

-1.46

 spons

-1.46

 tread

-1.45

 creep

-1.42

irts

-1.41

ospons

-1.41

POSITIVE LOGITS

Experts

1.60

Raw

1.53

Whoever

1.53

favorite

1.50

ashington

1.50

 Forensic

1.49

Ass

1.46

database

1.44

Jose

1.42

Coun

1.42

Activations Density 0.000%

the letter 'e'

No Comments

No Known Activations