INDEX

Explanations

past tense verbs and actions related to aggressive or violent incidents

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

0:0.02

1:0.02

2:0.06

3:0.06

4:0.12

5:0.03

6:0.04

7:0.38

8:0.05

9:0.03

10:0.08

11:0.06

Negative Logits

DragonMagazine

-2.00

 dermat

-1.49

igmatic

-1.47

earance

-1.46

��

-1.44

SEA

-1.42

 fuss

-1.41

rared

-1.40

icter

-1.40

edom

-1.39

POSITIVE LOGITS

ivated

1.84

forward

1.55

 throats

1.53

anium

1.53

 caches

1.50

 stairs

1.45

 catapult

1.44

 relentlessly

1.44

 chests

1.43

 Rapids

1.40

Activations Density 0.000%

past tense verbs and actions related to aggressive or violent incidents

No Comments

No Known Activations