INDEX

Explanations

quantitative descriptions related to counts or measurements

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.03

1:0.01

2:0.05

3:0.05

4:0.12

5:0.02

6:0.07

7:0.31

8:0.03

9:0.03

10:0.15

11:0.06

Negative Logits

hower

-1.78

nown

-1.51

nels

-1.40

ship

-1.36

cock

-1.35

holm

-1.35

 capsule

-1.35

netflix

-1.35

 Gleaming

-1.34

 encoded

-1.34

POSITIVE LOGITS

 bench

1.84

 dissent

1.71

 Judges

1.61

 benches

1.61

 dissenting

1.60

 Cato

1.56

 Appeals

1.53

Judge

1.52

Arg

1.51

 Prosecut

1.49

Activations Density 0.001%

quantitative descriptions related to counts or measurements

No Comments

No Known Activations