INDEX

Explanations

reasons or conditions for actions

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

1:0.03

2:0.36

3:0.07

4:0.07

5:0.04

6:0.06

7:0.08

8:0.05

9:0.03

10:0.05

11:0.08

Negative Logits

 synopsis

-1.85

ipel

-1.77

 unfocusedRange

-1.65

 Gloss

-1.63

 keywords

-1.58

DIR

-1.56

srf

-1.55

channelAvailability

-1.53

/,

-1.49

LIM

-1.48

POSITIVE LOGITS

smanship

1.55

 Cham

1.51

console

1.50

interstitial

1.46

AAA

1.44

imb

1.40

�

1.39

buy

1.39

put

1.39

 rise

1.36

Activations Density 0.001%

reasons or conditions for actions

No Comments

No Known Activations