INDEX

Explanations

actions related to laughing and humor

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L1_Hcat_z_lr1.20e-03_l18.00e-01_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v5.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.1.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

0:0.31

1:0.11

2:0.04

3:0.02

4:0.09

5:0.04

6:0.10

7:0.06

8:0.03

9:0.02

10:0.02

11:0.09

Negative Logits

 federally

-1.80

 contiguous

-1.69

 contingency

-1.68

 subsidized

-1.63

ensable

-1.60

 viable

-1.59

isible

-1.58

 contested

-1.56

 provisional

-1.54

 financed

-1.53

POSITIVE LOGITS

puff

1.96

lee

1.83

Redditor

1.69

Raven

1.66

roo

1.65

Pad

1.61

Mom

1.60

 Kramer

1.57

luck

1.54

Sn

1.49

Activations Density 0.039%

actions related to laughing and humor

No Comments

No Known Activations