INDEX

Explanations

quantitative metrics related to sports and economics

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

0:0.03

1:0.01

2:0.08

3:0.04

4:0.23

5:0.02

6:0.05

7:0.33

8:0.03

9:0.03

10:0.04

11:0.05

Negative Logits

izont

-1.74

ultz

-1.70

rame

-1.70

hene

-1.68

older

-1.66

eria

-1.65

ourage

-1.62

ura

-1.58

rification

-1.58

ador

-1.57

POSITIVE LOGITS

 confessions

1.56

 Travels

1.48

 Supplementary

1.48

 subsistence

1.44

 extras

1.40

 euth

1.37

 equivalent

1.36

 probabilities

1.35

 lecture

1.35

 Emin

1.34

Activations Density 0.001%

quantitative metrics related to sports and economics

No Comments

No Known Activations